Safety Evaluation in the Development of Medical Devices and Combination Products

about the book… Capturing the growth of the global medical device market in recent years, this practical new guide is essential for all who are responsible for ensuring safety in the use and manufacture of medical devices. It has been extensively updated to reflect significant advances, incorporating combination products and helpful case examples of current real-life problems in the field. The Third Edition explores these key current trends: • global device markets • continually advancing technology • the increasing harmonization of device safety regulation worldwide Each aspect of safety evaluation is considered in terms of International Standards Organization (ISO), US Food and Drug Administration (FDA), European Union (EU), and Japanese Ministry of Health and Welfare (MHW) perspectives. In addition, the book reflects the role of the continuing growth of technology in the incorporation of science, particularly in the areas of immunotoxicology and toxicokinetics. about the authors... SHAYNE C. GAD is the principal of Gad Consulting Services, Cary, North Carolina, USA. Gad Consulting Services is a 15-year-old consulting firm serving more than 350 clients (including 120 pharmaceutical companies in the US and 35 overseas). Dr. Gad received his B.S. in chemistry and biology from Whittier College, Whittier, California, and his Ph.D. in pharmacology and toxicology from the University of Texas at Austin, Austin, Texas, USA. He has published 39 books and more than 350 chapters, articles, and abstracts in the fields of toxicology, statistics, pharmacology, drug development, and safety assessment. Dr. Gad has more than 30 years of broad-based experience in toxicology, drug and device development, statistics, and risk assessment with specific expertise in neurotoxicology, in vitro methods, cardiovascular toxicology, inhalation toxicology, immunotoxicology, and genotoxicology. He was the past president of the American College of Toxicology and the Roundtable of Toxicology Consultants. Dr. Gad has consulted for the US Food and Drug Administration, Environmental Protection Agency, and the National Institutes of Health. MARIAN G. MCCORD is Associate Professor with appointments in the Department of Textile Engineering, Chemistry, and Science at North Carolina State University, Raleigh, NC, USA, the Joint UNC/NCSU Department of Biomedical Engineering in Raleigh and Chapel Hill, NC, USA, and an Adjunct Associate Professor of Medicine at the University of North Carolina at Chapel Hill. She serves as co-Director of the Atmospheric Plasma Laboratory at the College of Textiles at NCSU, and is a leader in development of Advanced Medical Technology at NCSU (www.ncsu.edu/amt). Dr. McCord received her M.S. in Bioengineering and Ph.D. in Textiles and Polymer Science from Clemson University, Clemson, South Carolina, USA. She has been active in several professional organizations, including the AAMI Protective Barriers Committee, ASTM F23, the Biomedical Engineering Society, the American Society for Engineering Education, and the Society for Biomaterials. Dr. McCord is the recipient of the 2007 Richard Felder Award for Outstanding Service in Teaching and Learning. She has been an invited lecturer and presenter at numerous consortia, meetings, and conferences, and has been a valued mentor and advisor for her students. Printed in the United States of America

•

Gad

McCord

71645

Safety Evaluation in the Development of Medical Devices and Combination Products, Third Edition

Toxicology

Safety Evaluation in the Development of Medical Devices and Combination Products Third Edition

Shayne C. Gad Marian G. McCord



Safety Evaluation in the Development of Medical Devices and Combination Products

Safety Evaluation in the Development of Medical Devices and Combination Products Third Edition

Shayne C. Gad

Gad Consulting Services Cary, North Carolina, USA

Marian G. McCord

North Carolina State University Raleigh, North Carolina, USA

Informa Healthcare USA, Inc. 52 Vanderbilt Avenue New York, NY 10017 © 2008 by Informa Healthcare USA, Inc. Informa Healthcare is an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 1-4200-7164-5 (Hardcover) International Standard Book Number-13: 978-1-4200-7164-1 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Library of Congress Cataloging-in-Publication Data Gad, Shayne C., 1948Safety evaluation in the development of medical devices and combination products / by Shayne C. Gad, Marian G. McCord. p. ; cm. Rev. ed. of: Safety evaluation of medical devices / Shayne Cox Gad. 2nd ed., rev. and expanded. c2002. Includes bibliographical references and index. ISBN-13: 978-1-4200-7164-1 (hardcover : alk. paper) ISBN-10: 1-4200-7164-5 (hardcover : alk. paper) 1. Biomedical materials--Testing. 2. Biomedical materials--Toxicity testing. 3. Biocompatibility--Testing. I. McCord, Marian G. II. Gad, Shayne C., 1948- Safety evaluation of medical devices. III. Title. [DNLM: 1. Biocompatible Materials. 2. Materials Testing. 3. Equipment Failure. 4. Equipment Safety. 5. Toxicity Tests. QT 37 G12299s 2008] R857.M3G33 2008 610.28'4--dc22 2008027552 Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. For Corporate Sales and Reprint Premissions call 212-520-2700 or write to: Sales Department, 52 Vanderbilt Avenue, 7th floor, New York, NY 10017. Visit the Informa Web site at www.informa.com and the Informa Healthcare Web site at www.informahealthcare.com

To Joyce, my wife, partner, and best friend, for all the light and joy that she has brought into my world. Shayne Cox Gad To Mickey with love—you are the compass that guides me, the ballast that keeps me afloat, and the wind in my sails. Marian McCord

Preface to the Third Edition

This third edition of Safety Evaluation of Medical Devices, while continuing to focus on the objective of the earlier editions (to serve as a single-volume practical guide for those who are responsible for or concerned with ensuring safety in the use and manufacture of medical devices) also reflects the significant changes brought about since the last (second) edition some six years ago. It not only updates throughout, but also adds extensive coverage of combination products and case histories of current real life problems in the field. Foremost, this new edition has been recast throughout to address the fact that device markets are global, that technology continues to advance, and that device safety regulation has been increasingly harmonized. Each aspect of device safety evaluation is considered in terms of International Standards Organization (ISO), U.S. Food and Drug Administration (FDA), European Union (EU), and Japanese Ministry of Health and Welfare (MHW) perspectives. Additionally, the continuing growth of technology has led to the incorporation of science (particularly in the areas of immunotoxicology and toxicokinetics). Also incorporated are new case examples and citations with the means of access to Internet-based regulatory and scientific sites, reflecting the universal adoption of this technology into our world. Shayne C. Gad Marian G. McCord

Contents

Preface / v Table of Contents

Conclusions / 67 References / 68

/ vii

1. Introduction / 1

6. Cytotoxicity Testing / 69

Biocompatibility / 1 Scope of Devices and the Medical Device Market / 2 History / 3 Nonspecific Regulatory Considerations / 4 Organizations Regulating Drug and Device Safety in the United States / 5 References / 6

Background / 69 Crystal Violet Staining / 70 Silicone Microphysiometer / 70 Microtox Test / 70 Neutral Red Uptake Assay / 70 Agar Diffusion Test / 72 Direct Contact Test / 72 Elution Test / 73 Correlation with in vivo Results / 73 Conclusion / 73 References / 73

2. Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation / 7 Regulatory Basis / 7 Toxicity Testing: Medical Devices / 11 References / 26

3. Road Map to Test Selections

7. Hemocompatibility / 75 Non-Contact Devices / 75 External Communicating Devices / 75 Implant Devices / 76 Standard Tests / 77 Platelets / 81 Conclusion / 82 References / 82

/ 27

Key Concepts / 27 Condition of Use / 27 Materials, Components, and Products / 27 Chemical and Physical Property Considerations / 27 Specific Material Considerations / 30 Physical Properties of Polymers, Elastomers, and Silicones / 37 Biologically Derived Materials / 38 Factors Influencing Test Selection / 38 Prior Knowledge / 40 Miscellaneous Reference Sources / 41 Search Procedure / 42 Monitoring Published Literature and other Research in Progress / 42 New Sources / 42 Kinds of Information / 43 Types and Uses of Tests / 45 Reasonable Man / 45 Qualifications Versus Process Control / 45 Tiers of Concern: Consumers, Health Care Providers, and Manufacturing Employees / 45 Sterilization and Cleanliness / 45 References / 46

4. Materials in Medical Device Design Introduction / 49 Metals / 49 Ceramics / 51 Polymers / 52 Elastomers / 54 Tissue Engineering Scaffolds Nanomaterials / 59 References / 60

8. Local Tissue Tolerance / 83 Dermal Irritation / 83 Ocular Irritation Testing / 86 Other Nonparenteral Route Irritation Tests Parenteral Irritation/Tolerance / 89 Parenteral Routes / 89 Intracutaneous Irritation / 91 Factors Affecting Irritation Responses and Test Outcome / 95 References / 96

/ 89

9. Immunotoxicology / 99 Overview of the Immune System / 100 Immunotoxic Effects / 102 Immunosuppression / 103 Immunostimulation / 103 Hypersensitivity / 104 Evaluation of the Immune System / 107 Immunopathologic Assessments / 108 Local Lymph Node Assay (LLNA) / 124 Approaches / 126 Problems and Future Directions / 127 References / 128

/ 49

/ 59

10. Implantation Biology and Studies

/ 133

USP Implantation Test / 133 British Pharmacopoeia / 133 ISO 10993 Implantation Test / 134 Test Method for Implantation in Subcutaneous Tissue / 136 Test Method for Implantation in Muscle / 137 Test Method for Implantation in Bone / 137 Control Materials / 138 Long-Term Implant Studies / 138

5. What to Test: Sampling and Sample Preparation / 63 Sampling / 63 Randomization / 64 Sample Preparation / 64 Reference Materials / 67

vii

viii

Contents

17. Clinical Studies for Medical Devices / 231

Considerations / 141 References / 142

11. Acute Systemic Toxicity Testing Device Safety Evaluation / 143 Introduction / 143 Acute Systemic Toxicity Characterization References / 149

/ 143

12. Genotoxicity / 151 DNA Structure / 151 Cytogenetics / 158 Plate Incorporation Assay / 163 In vitro Cytogenetic Assays / 171 In vivo Cytogenetic Assays / 174 References / 176

13. Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity / 181 Objectives / 181 Regulatory Considerations / 181 Study Design and Conduct / 182 Parameters to Measure / 182 Histopathology / 184 Study Interpretation and Reporting / 186 Reproductive and Developmental Toxicity / 186 Dose and Sample Preparation / 188 Methodological Issues / 188 Gravid Uterine Weights / 189 Implant Counts and Determination of Pregnancy / 189 Developmental Signs / 190 Data Interpretation / 191 References / 194

14. Carcinogenicity / 197 Animal Model / 197 Dose Selection / 198 Interpretation of Results References / 201

/ 201

15. Degradation Products and Impurities Introduction / 203 Metals / 203 Ceramics and Glasses / 206 Polymers / 208 Effects of Sterilization / 211 Heavy Metals / 215 Appendix A: Extractables Study / 216 References / 220

16. Combination Products / 223 Historical Background / 223 Future Trends / 223 References / 229

/ 203

Design Considerations / 231 Epidemiology / 237 Conclusion / 240 References / 241

18. Special Studies / 243 Cardiovascular Devices and Prostheses / 243 Tampons / 245 Contact Lenses and Their Solutions / 245 Specific Tests / 245 21-Day Eye Irritation Study in Rabbits / 250 Systemic Injection Test / 251 Production of Toxic Shock Syndrome Toxin-1 (TSST-1) / 251 In Vitro Evaluation of the Effects of Potential Tampon Materials on pH and Growth of Vaginal Microflora / 252 Inflammatory Responses to Biomaterials / 252 References / 253

19. Toxicokinetics in Biomaterial and Device Safety Evaluation / 255 Introduction / 255 Regulations / 256 Principles / 256 Laboratory Methods / 265 Physiologically-based Pharmacokinetic (PBK) Modeling / 272 Biologically Derived Materials / 273 References / 273

20. Case Histories and Problem Resolution

/ 275

Phthalate and Other Residuals Leaching from Devices / 275 Dalkon Shield / 276 Talc on Gloves and Condoms / 277 Toxic Shock Syndrome / 277 Latex Allergy / 279 Silicones in Devices / 280 Problem Solving / 282 Risk Assessments / 283 Case Studies / 284 Conclusions / 286 References / 286

Appendix A: Selected Regulatory and Toxicological Acronyms and Abbreviations / 289 Appendix B: Contract Testing Laboratories

/

291

Appendix C: Notable International Regulatory Internet Addresses / 293 Appendix D: Non-U.S. Medical Device Regulators / 295 Index / 297

1 Introduction

The medical device industry in the United States and worldwide is immense in its economic impact (sales in 2006 were $220 billion worldwide and $114 billion in the United States, $60 billion in the European Community and $32 billion in Japan), scope (between 92,000 and 145,000 different devices are produced in the U.S. by ~12,000 different manufacturers employing some 370,000 people; it is believed that ~2,100 of these manufacturers are development stage companies without products yet on the market) and importance to the health of the world’s citizens (The Wilkerson Group, 2006). The assessment of the safety to patients using the multitude of items produced by this industry is dependent on schemes and methods which are largely particular to these kinds of products, not as rigorous as those employed for foods, drugs, and pesticides, and which are in a state of flux. Regulation of such devices is, in fact, relatively new. It is only with the Medical Device Amendments (to the Food, Drug and Cosmetic Act) of 1976 that devices have come to be explicitly regulated at all. And with the Safe Medical Devices Act of 1990, the Medical Device Amendments of 1992 and subsequent laws that the regulation of devices for biocompatibility became rigorous. The causes behind this timing are reviewed in the case histories presented in the last chapter of this book. For purposes of this book, the safety we are concerned with is that related to the biological and chemical interactions of devices with patients’ bodies, and not that due to mechanical or structural malfunction (such as structural failure of heart valves and pacemakers). Such safety, also referred to as biocompatibility, only became of general concern to the public with publicity around plasticizers in devices and increased mortality with cardiovascular stents. Earlier cases of perceived significant risk on the part of devices (the Dalkon Shield intrauterine device, silicones in breast implants, and latex present in gloves and a wide range of other devices) largely faded from public and professional memory by the beginning of the 21st century.

required data and testing is expanded, what constitutes adequate testing is a moving target as time passes. There is no universally accepted definition for biomaterial and biocompatibility. Yet the manufacturer who ultimately markets a device will be required by FDA to demonstrate biocompatibility of the product as part of the assurance of its safety and effectiveness. The manufacturer is responsible for understanding biocompatibility tests and select methods which best demonstrate • the lack of adverse biological response from the biomaterial, and • the absence of adverse effects on patients. Diversity of the materials used, types of medical devices, intended uses, exposures, and potential harms present an enormous challenge to design and conduct welldefined biocompatibility testing programs. Experience gained in one application area is not necessarily transferable to another application. The same applies to different or sometimes slightly different (variable) materials. Biodegradation and interaction of materials complicates safety considerations, as does the increased scope of combination device drug products. Biocompatibility describes the state of a biomaterial within a physiological environment without the material adversely affecting the tissue or the tissue adversely affecting the material. Biocompatibility is a chemical and physical interaction between the material and the tissue and the biological response to these reactions. Biocompatibility assays are used to predict and prevent adverse reactions and establish the absence of any harmful effects of the material. Such assays help to determine the potential risk which the material may pose to the patient. The proper use of biocompatibility tests can reject potentially harmful materials while permitting safe materials to be used for manufacturing the device. Any biocompatibility statement is useful only when it is considered in the proper context. A statement such as “propylene is biocompatible” lacks precision and can lead to misunderstanding. Any statement of biocompatibility should include information on the type of device, intended conditions of use, degree of patient contact, and the potential of the device to cause harm. Manufacturers should avoid using the term “biocompatible” without clearly identifying the environment in which it is used and any limitations on such. The need for biocompatibility testing and the extent of such testing that should be performed depends on numerous factors which are presented and considered in Chapter 2. These factors include the type of device, intended use, liability, degree of patient contact, nature of the components, and potential of the device to cause harm. There are no universal tests to satisfy all situations, and there is no single test which can predict biological performance of the material

BIOCOMPATIBILITY A medical device that is adequately designed for its intended use should be safe for that use. The device should not release any harmful substances into the patient which can lead to adverse effects. Some manufacturers believe that biocompatibility is sufficiently indicated if their devices are made of medical grade material, or materials approved by FDA as direct or indirect additives. The term “medical grade” does not have an accepted legal or regulatory definition and can be misleading without biocompatibility testing. Likewise, the existence of a material master file (MMF) does not provide any assurance as to what biocompatibility data (or of what quality) is available. More to the point as the extent of 1

2

Safety Evaluation of Medical Devices

or device and reliably predict the safety of the device. The types and intended uses of medical devices determine the types and number of tests required to establish biocompatibility. Biological tests should be performed under the condition which simulates the actual use of the product or material as closely as possible and should demonstrate the biocompatibility of a material or device for a specific intended use. These tests will be more extensive for a new material than for those materials that have an established history of long and safe uses. All materials used in the manufacture of a medical device should be considered for evaluation of their suitability for intended use. Consideration should always be given to the possibility of the release of toxic substances from the base materials, as well as any contaminants which might remain after the manufacturing process or sterilization. The extent of these investigations will vary depending on previously known information (prior art) and initial screening tests.

Fundamentals of Biocompatibility Tests Biocompatibility is generally demonstrated by tests utilizing toxicological principles which provide information on the potential toxicity of materials in the clinical application. Many classical toxicological tests, however, were developed for a pure chemical agent, and are not applicable to biocompatibility testing of materials. In addition, medical devices are an unusual test subject in toxicity testing. A biomaterial is a complex entity, and the material toxicity is mediated by both physical and chemical properties. Toxicity from biomaterial often comes from leachable components, and the chemical composition of a material is often not known. Toxicological information on the material and its chemical composition is seldom available, and the possible interactions among the components in any given biological test system are seldom known. Biocompatibility cannot be defined by any single test. It is highly unlikely that any single parameter will be able to ensure biocompatibility. Therefore, it is necessary to test as many biocompatibility parameters as appropriate to develop a matrix of information. It is also important to test as many samples as possible. Therefore, suitable positive and negative controls should produce a standard response index for repeated tests. Additionally, the use of exaggerated conditions, such as using higher dose ranges and longer contact durations or multiple insults that are may factors more severe than the actual use condition, is important. Identifying and subsequently ensuring an acceptable exposure level that is multiple factors below the lowest toxic level is the general and expected practice. Most of the basic biocompatibility tests are short-term tests to establish acute or short-term toxicity. Data from these short-term tests should not be stretched to cover the areas where no test results are available. Biocompatibility testing should be designed to assess the potential adverse effects under actual use conditions or specific conditions close to the actual use conditions. The physical and biological data obtained from biocompatibility tests should be correlated to the device and its use. Accuracy, reproducibility, and interpretability of tests depend on the method and equipment used and the investigator’s skill and experience. There are several toxicological principles which the investigator must consider before planning biocompatibility

testing programs. Biocompatibility depends on the tissue that contacts the device. For example, the requirements for blood-contacting device would be different from those applicable to a urethral catheter. Also, the degree and nature of required biocompatibility assurance depends on the nature, extent and duration of contact with the human body. Some materials, such as those used in orthopedic implants, are meant to last for a long period in the patient. In this case, a biocompatibility testing program needs to show that the implant does not adversely affect the body during the long period of use. The possibility of biodegradation of material or device can not be ignored, and evaluation of such is now required by ISO-10993 guidances.. Biodegradation by the body can change an implant’s safety and effectiveness. The leachables from plastic used during a hemodialysis procedure may be very low, but the patient who is dialyzed three times a week may be exposed to a total of several grams during their lifetime. Therefore, cumulative effects (chronicity) should be assessed. Two materials having the same chemical composition but different physical characteristics may not induce the same biological response. The nature of the tissue to device interface (is the device surface smooth textured or rough?) is very important. Also, past biological experiences with seemingly identical materials also has possible limited toxicity. Toxicity can arise from leachable components of the material due to differences in formulation and manufacturing procedures. Empirical correlation between biocompatibility testing results and actual toxicity findings in humans and the extrapolation of the quantitative results from short-term in vitro tests to quantitate toxicity at the time of use are controversial. These need careful and scientifically sound interpretation and adjustment. The control of variation in biological susceptibility and resistance to obtain a biological response range for toxic effect, and host factors which determine the variability of susceptibility in toxicological response adjustment to susceptibility in the human population also need careful attention. The challenge of biocompatibility is to create and use knowledge to reduce the degree of unknowns and to help make the best possible decisions. The hazard presented by a substance, with its inherent toxic potential, can only be manifested when fully exposed in a patient. Therefore, risk, which is actual or potential harm, is a function of toxic hazard and exposure. The safety of any leachables contained in the device or on the surface can be evaluated by determining the total amount of potentially harmful substance, estimating the amount reaching the patient tissues, assessing the risk of exposure, and performing the risk versus benefit analysis. When the potential harm from the use of biomaterial is identified from the biocompatibility tests, this potential must be compared against the availability of an alternate material.

SCOPE OF DEVICES AND THE MEDICAL DEVICE MARKET According to section 201(h) of the Food, Drug and Cosmetic Act, a medical device is defined as an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including a component, part, or accessory that is: • Recognized in the official National Formulary, or the United States Pharmacopoeia (USP, 1994), or any supplement to them.

Chapter 1:

Intended for use in the diagnosis of disease, in man or other animals, or • Intended to affect the structure or any function of the body or man or other animals, and which does not achieve any of its primary intended purposes through chemical action within or on the body of man or other animals, and which is not dependent upon being metabolized for the achievement of any of its principal intended purposes (CDRH, 1992). Under this definition, devices might be considered as belonging to one of eight categories (North American Industrial Classification): surgical and medical instruments, ophthalmic, dental, laboratory apparatus, irradiation, specialty devices, medical/surgical supplies, in vitro diagnostics, and electromedical. There are (in 2000) 16,170 companies involved in these sectors—6750 of them manufacture worldwide. This is a global industry with a $220 billion annual market. The US market is $114 billion, or 52% of this (MDDI, 2000). The top twenty medical devices by revenues in 1999 were: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Incontinence supplies Home blood glucose monitoring products Wound closure products Implantable defibrillators Soft contact lenses Orthopedic fixation devices Pacemakers Examination gloves Interventional cardiovascular coronary stents Arthroscopic accessory instruments Prosthetic knee joint implants Lens care products Prosthetic hip joint implants Multiparameter patient monitoring equipment Mechanical wound closure Wound suture products Absorbable polymers Hearing aids Wheelchair and scooter/mobility aids industry Peritoneal dialysis sets.

The ten projected biggest growth device products (in 2000) are shown in Table 1.1.

HISTORY As has previously been reviewed by Hutt (1989), the regulation of medical devices has followed a different history than

Table 1.1

Introduction

that of drugs. Medical devices go back to at least the Egyptians and Etruscans. Problems with fraudulent devices in the United States date back to the late 1700s, though no legislative remedy was attempted until the 1900s. In fact, the legislative history of the 1906 Food and Drug Act contains no references to devices. Devices continued to be regulated under the postal fraud statutes. Such regulation was evidently ineffectual, as fraudulent devices flourished during this period. Starting in 1926, the Food and Drug Administration (FDA) monitored such devices and assisted the U.S. Postal service in its regulatory actions. Medical devices were covered in the 1938 Act, but only in regard to adulteration and misbranding. Over the intervening years, various committees which examined medical device regulation consistently came to similar conclusions: that the FDA has inadequate authority and resources to regulate the medical device industry. As part of the agreement that resulted in passage of the 1962 amendments, however, all references to medical devices were deleted. The need and demand for increased regulation continued to grow. In 1967, President Lyndon Johnson supported the proposed Medical Device Safety Act, which nevertheless was not well received by Congress. In fact, no legislation pertaining to medical device safety was passed until 1976. In 1969, at the request of then President Richard Nixon, the Department of Health, Education and Welfare (HEW) established a Study Group in Medical Devices, also known as the Cooper Committee, because it was chaired by the Director of the National Heart and Lung Institute, Dr. Theodore Cooper. Its report in 1970 concluded that a different regulatory approach was needed to deal with medical devices. This report initiated the chain of events that culminated in the Medical Device Amendment of 1976. In the interim, the Bureau of Medical Devices and Diagnostic Products was created in 1979. Remarkably, the 1976 Amendment retained the essential provisions of the Cooper Committee Report regarding inventory and classification of all medical devices by class: Class I (general controls), Class II (performance standards), or Class III (premarket approval). These classifications are discussed in greater detail later in this chapter. These remain the essential regulations applicable to medical devices. Both the Drug Price Competition and Patent Restoration Act of 1984 and the Orphan Drug Act of 1983 contained language that made the provisions of the laws applicable to medical devices but did not have provisions unique to medical devices. The recent perceptions, revelations, and controversy surrounding silicone breast implants will probably cause additional changes in the regulation of devices.

Highest Revenue Growth Products

Rank

Product

1 2 3 4 5 6 7 8 9 10

Fibrin sealants Solid artificial organs Left ventricular assist devices Skin substitute products Refractive surgical devices Gynecologic falloposcopes PTMR products Bone growth substitutes and growth factors Growth factor dressings Vascular stent-grafts

Source: Frost and Sullivan.

3

Revenue Growth Rate (%) (yrs) 174.6 (95-02) 141.2 (95-02) 96.0 (95-02) 63.1 (97-04) 54.4 (98-05) 49.5 (95-00) 47.8 (00-04) 47.0 (97-04) 46.0 (97-04) 46.0 (97-04)

Specialty Wound care Transplant/implant Cardiovascular Wound care Ophthalmic Endoscopic/MIS Cardiovascular Orthopedics Wound care Cardiovascular

4

Safety Evaluation of Medical Devices

As a consequence, 1978 brought guidelines for investigational device exemptions (IDEs, the equivalent of INDAs for drugs). These requirements, as shall be seen later, effectively excluded a wide range of medical devices from regulation by establishing an exemption for those new or modified devices which are equivalent to existing devices. The year 1990 saw the passage of the Safe Medical Devices Act, which made premarketing requirements and postmarketing surveillance more rigorous. The actual current guidelines for testing started with the USP guidance on biocompatibility of plastics. A formal regulatory approach springs from the Tripartite agreement, which is a joint intergovernmental agreement between the United Kingdom, Canada, and the United States (with France having joined later). After lengthy consideration, the FDA has announced acceptance of International Standards Organization (ISO) 10993 guidelines for testing (ASTM, 1990; FAO, 1991; MAPI, 1992; O’Grady, 1990; Spizizen, 1992) under the rubric of harmonization. This is the second major trend operative in device regulation: the internationalization of the market place with accompanying efforts to harmonize regulations. Under ICH (International Conference on Harmonization) great strides have been made in this area. Independent of FDA initiatives, the USP has promulgated test methods and standards for various aspects of establishing the safety of drugs (such as the recent standards for inclusion of volatiles in formulated drug products), which were, in effect, regulations affecting the safety of drugs and devices. Most of the actual current guidelines for the conduct of nonclinical safety evaluations of medical devices have evolved from such quasi-agency actions (such as the USP’s 1965 promulgation of biological tests for plastics and ongoing American National Standards Institute (ANSI) standard promulgation). Public concerns about three specific device safety issues have seemed to increase regulatory scrutiny. The first of these, the Dalkon Shield, was an intrauterine contraceptive device produced by the A. H. Robbins Corporation (Sivin, 1993). Its use was associated with unacceptable rates of pregnancy, pelvic inflammatory disease, and death in women who used it. The device was withdrawn from the market in 1974, and in 1988 Robbins reached a $3.3 billion settlement in response to a class action suit (Nocera, 1995). The second case is that of silicone-filled breast implants, which have been purported to cause a range of autoimmune and neurologic effects on some women who have them. Though the validity of these claims remains unproven or disproven, litigation over them drove the primary manufacturer (Dow Corning) into bankruptcy and lead to the removal of these products from the market (though, in 2006, they have returned to the market). Since the late 1980s concern has grown about allergic responses to latex in devices. Several deaths have been blamed on anaphylactic responses to such effects (Lang, 1996). In the current century, potential male reproductive effects from DEHP leaching from medical devices and leading to removal of such products from the market place.

NONSPECIFIC REGULATORY CONSIDERATIONS A broad scope review of regulatory toxicology is presented in Gad (2001). Some necessity to understand regulations beyond those covered in Chapter 2 requires review here, however.

Good Laboratory Practices The original promulgation of GLPs was by the U.S. FDA in 1978 in response to a variety of cases which led the agency to conclude that some of the data that it had obtained in support of product approvals were not trustworthy. Subsequently, other regulatory agencies and authorities in the United States and across the world have either promulgated their own version of similar regulations or required adherence to the set generated by the U.S. FDA or another body. The EEC requirement for compliance with GLPs for safety tests has recently been reinforced in a modification of Directive 75/318/EEC (Regulatory Affairs Focus, 1996; ISO, 1990; European Committee for Standardization, 1991). The FDA last revised the GLP regulations in 1989 (FDA, 1989), but is currently (July, 2008) working on a revision. The GLPs require that all pivotal preclinical safety studies—that is, those that are used and regulatorily required to make decisions as to the safety of the product (in our case, a device)—conducted under a well-defined protocol utilizing procedures set forth in written standard operating procedures by trained (as established by documentation) personnel under the direction of a study director. All work must be reviewed by an independent Quality Assurance Unit (QAU). The regulations require rigorous attention to record keeping, but do not dictate how actual studies are designed or conducted in a technical sense (Gad and Taulbee, 1996).

Animal Welfare Act (AWA) Gone are the days when the biomedical research scientist could conduct whatever procedures or studies that were desired using experimental animals. The Animal Welfare Act (APHIS, 1989) (and its analogues in other countries) rightfully requires careful consideration of animal usage to ensure that research and testing uses as few animals as possible in as humane a manner as possible. As a start, all protocols must be reviewed and approved by an Institutional Animal Care and Use Committee (IACUC) prior to animals being ordered or a study being initiated. Such review takes time, but should not serve to hinder good science. When designing a study or developing a new procedure or technique, the following points should be kept in mind: 1. Will the number of animals used be sufficient to provide the required data, yet not constitute excessive use? It ultimately does not reduce animal use to utilize too few animals to begin with and then have to repeat the study. 2. Are the procedures employed the least invasive and traumatic available? This practice is not only required by regulations, but is also sound scientific practice, since any induced stress will produce a range of responses in test animals that can mask or confound the chemically induced effects. Most recently (September of 2000) USDA (which administers the AWA) had decided to begin including rodents in all aspects of the AWA’s reporting requirements.

Regulations Versus Law A note of caution must be inserted here. The law (the document passed by Congress) and the regulations (the documents written by the regulatory authorities to enforce the laws) are separate documents. The sections in the law

Chapter 1:

do not necessarily have numerical correspondence. For example, the regulations on the PMA process are described in 21 CFR 312, but the law describing the requirement for a PMA process is in Section 515 of the FDLI. Because the regulations rather than the laws themselves have a greater impact on toxicological practice, greater emphasis is placed on regulation in this chapter. For a complete review of FDA law, the reader is referred to the monographs by Food and Drug Law Institute in 1995 (FDLI, 1995). Laws authorize the activities and responsibilities of the various federal agencies. All proposed laws before the U.S. Congress are referred to committees for review and approval. The committees responsible for FDA oversight are summarized in Table 1.2. This table also highlights the fact that authorizations and appropriations (the funding necessary to execute authorizations) are handled by different committees. Figure 1.1 presents the organization of the Center for Devices and Radiological Health (CDRH). As can be seen by the organizational structure presented in the figure, the categorization of devices for division review purposes is functionally based.

ORGANIZATIONS REGULATING DRUG AND DEVICE SAFETY IN THE UNITED STATES The agency formally charged with overseeing the safety of drugs and devices in the United States is the FDA. It is headed by a commissioner who reports to the Secretary of the Department of Health and Human Services (DHHS) and has a tremendous range of responsibilities. Medical devices are overseen by the CDRH, headed by a director. Drugs are overseen primarily by the Center for Drug Evaluation and

Table 1.2

Introduction

5

Congressional Committees Responsible for FDA Oversight

Authorization Senate

House

Appropriation Senate

House

All public health service agencies are under the jurisdiction of the Labor and Human Resources Committee. Most public health agencies are under the jurisdiction of the Health and the Environmental Subcommittee of the House Energy and Commerce Committee.

Unlike most other public health agencies, the FDA is under the jurisdiction of Agriculture, Rural Development, and Related Agencies Subcommittee of the Senate Appropriations Committee. Under the jurisdiction of the Agriculture, Rural Development, and Related Agencies Subcommittee of the House Appropriations Committee.

Research (CDER) (though some therapeutic or health care entities are considered as biologically derived and therefore regulated by the Center for Biologic Evaluation and Research, or CBER). There are also “combination products” (part drug, part device) which may be regulated by either or both CDER/CBER and CDRH, depending on the principal mode of action (PMOA) of the product. Most of the regulatory interaction of a toxicologist involved in assessing the biocompatibility of devices is with the appropriate part of the CDRH, though for combination products the two centers charged with drugs or biologicals may also come into play. Within the CDRH there is a range of groups (called divisions) which focus on specific areas of

DEPARTMENT OF HEALTH AND HUMAN SERVICES FOOD AND DRUG ADMINISTRATION CENTER FOR DEVICES AND RADIOLOGICAL HEALTH OFFICE OF THE CENTER DIRECTOR DIRECTOR Daniel G. Schultz, M.D. DEPUTY DIRECTOR Linda S. Kahan ASSOCIATE DIRECTOR FOR MANAGEMENT AND SYSTEMS AND OFFICE OF MANAGEMENT OPERATIONS DIRECTOR Ruth E. Mckee ( DBWA)

OFFICE OF COMPLIANCE

OFFICE OF DEVICE EVALUATION

DIRECTOR Timothy A. Ulatowski

OFFICE OF SCIENCE AND ENGINEERING LABORATORIES DIRECTOR Larry G. Kessler, Sc.D.

(DBWE)

Main Tel: 240-276-3939 Fax: 240-276-3943 Room 100E, Bldg CORP, 9200 Corporate BLVD Rockville, MD20850 MailCode: HFZ-1

(DBWB)

DIRECTOR Donna B. Tillman, Ph.D.

(DBWC)

OFFICE OF COMMUNICATION, EDUCATION, AND RADIATION PROGRAMS

OFFICE OF SURVEILLANCE AND BIOMETRICS

OFFICE OF IN VITRO DIAGNOSTIC DEVICE EVALUATION AND SAFETY

DIRECTOR Lynne L. Rice

DIRECTOR Susan N. Gardner, Ph.D.

DIRECTOR Steven I. Gutman, M.D.

(DBWG)

(DBWH)

(DBWJ)

Prepared by the Management Programs & Analysis Staff, OMP, OM-06-01-07

Figure 1.1 Organizational chart of the Office of Device Evaluation (ODE) for the Center for Devices and Radiological Health (CDRH) of the FDA. Current officials (as of 6/01/2007) are identified by name. ODE evaluates submissions for new device approvals. Source: http://www.fda.gov/oc/orgcharts/orgchart.html.

6

Safety Evaluation of Medical Devices

use for devices (such as general and restorative devices; cardiovascular, respiratory, and neurological devices; ophthalmic devices; reproductive, abdominal, ear, nose, and throat, and radiological devices; and clinical laboratory devices). Within each of these there are engineers, chemists, pharmacologists/toxicologists, statisticians, and clinicians. There is also at least one nongovernmental body which must review and approve various aspects of devices, setting forth significant “guidance” for the evaluation of safety of devices. This is the USP, and its responsibilities and guidelines are presented later in Chapter 2.

REFERENCES APHIS (1989). Animal and Plant Health Inspection Service, United States Department of Agriculture. Federal Register, 54(168), 36112–36163, August 31, 1989. ASTM (1990). Standardization in Europe: A success story. ASTM Standardization News, 38, February 1990. CDRH (1992). Regulatory Requirements for Medical Devices: a Workshop Manual. Center for Device and Radiological Health, HHS Publication FDA 92-4165, August 1992. European Committee for Standardization (1991). CEN Annual Report 1991. Brussels. FAO (1991). Report of the FAO/WHO Conference on Food Standards, Chemicals in Food and Food Trade (in cooperation with GATT), Vol. 1. Rome, March 18–27, 1991. FDLI (1995–1998). Compilation of Food and Drug Laws, Volumes I, II, III and Supplement. Food and Drug Law Institute, Washington, D.C. FDA (1989). Good Laboratory Practice Regulations: Final Rule. Federal Register, Part VI, Vol. 52, No. 172, September 4, 1989.

Gad, S.C. (2001). Regulatory Toxicology, 2nd edn. Taylor & Francis, Philadelphia, PA. Gad, S.C. and Taulbee, S. (1996). Handbook of Data Recording, Maintenance and Management for the Biomedical Sciences. CRC Press, Boca Raton, FL. Hutt, P.B. (1989). A history of government regulation and misbranding of medical devices. Food Drug Cosm. Law J., 44(2): 99–117. ISO (1990). ISO 9000 International Standards for Quality Management Vision 2000—A Strategy for International Standards’ Implementation in the Quality Arena During the 1990s (2nd edn., compendium). EEC, Brussels. Lang, L.A. (1996). A review of latex hypersensitivity. Toxic Subst. Mech., 15: 1–11. MAPI (1992). The European community’s new approach to regulation of product standards and quality assurance (ISO 9000): What it means for U.S. manufacturers. MAPI Economic Report ER-218, January 1992. MDDI (2000). Industry Snapshot, Medical Device and Diagnostics Industry. December, pp. 47–56. Nocera, J. (1995). Fatal litigation, Fortune, October 16, pp. 60–82. O’Grady, J. (1990). Interview with Charles M. Ludolph. ASTM Standardization News, 26, February, 1990. Regulatory Affairs Focus (1996). European update. Regulatory Affairs Focus, 1(4): 8. Sivin, I. (1993). Another look at the Dalkon Shield: Meta-analysis underscores the problems. Contraception, 48: 1–12. Spizizen, G. (1992). The ISO 9000 standards: Creating a level playing field for international quality. National Productivity Review, Summer, 1992. USP (1994). The United States Pharmacopoeia, XXIII revision. United States Pharmacopoeial Convention, Rockville, MD. The Wilkerson Group (2006). Forces Reshaping the Performance and Contribution of the U.S. Medical Device Industry. Health Industry Manufacturers Association, Washington, D.C.

2 Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

As discussed in Chapter 1, in the United States, according to 201(h) of the Food, Drug and Cosmetic Act, a medical device is defined as an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including a component, part, or accessory that is:

on the day the amendment was passed were assigned to one of three classes (I, II, or III), based on the recommendation of advisory panels. Medical device classification procedure is described in Part 860. Class I products (the least risk burdened) were those for which safety and effectiveness could be reasonably assured by general controls. Such devices are available over the counter to the general public. Class II products were those for which a combination of general controls and performance standards were required to reasonably assure safety and effectiveness. Class II devices are generally available only with a doctor’s prescription, but may be used at home. Class III products were those for which general controls and performance standards were inadequate; these were required to go through a premarket approval process. All devices commercially distributed after May 28, 1976 (“preamendment Class III devices”) which are not determined to be substantially equivalent to an existing marketed device are automatically categorized as Class III and require the submission of a PMA. Please note that these are classifications for regulatory purposes only and are distinct from the classification (HIMA/ PHRMA) of product types (e.g., internal versus external) discussed elsewhere in this chapter. Kahan (1995) provides a detailed overview of what comprises general controls, performance standards and such. As with the subchapter on drugs, much of the subchapter on medical devices in the regulations concerns categorizations and specifics for a wide variety of devices. For a toxicologist involved in new product development, the parts of highest interest are 812 and 814. As with drugs, devices must be shown to be safe and effective when used as intended, and data must be provided to demonstrate such claims. In order to conduct the appropriate clinical research to obtain these data, a sponsor applies to the Agency for an IDE, as described in 21 CFR 812. As stated in this section, “an approved investigational device exemption (IDE) permits a device that would otherwise be required to comply with a performance standard or to have premarket approval to be shipped lawfully for the purpose of conducting investigations of that device.” Given the broad range of products that fall under the category of medical devices, the toxicological concerns are equally broad; testing requirements to support an IDE are vaguely mentioned in the law, even by FDA standards. In this regard, the law simply requires that the IDE application must include a report of prior investigations which “shall include reports of all prior clinical, animal and laboratory testing.” There is no absolute written requirement for animal testing, only a requirement that such testing must be reported. There are, of course, standards and conventions to be followed in designing a safety package to support an IDE, and these are discussed in a subsequent section of this chapter.

(a) recognized in the official National Formulary, or the United States Pharmacopoeia (USP, 2007), or any supplement to them, and (b) intended for use in the diagnosis of disease or other condition, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals, or intended to affect the structure or any function of the body or man or other animals, and which does not achieve any of its primary intended purposes through chemical action within or on the body of man or other animals, and which is not dependent upon being metabolized for the achievement of any of its principal intended purposes (CDRH, 1992).

REGULATORY BASIS Regulations: General Considerations for United States The U.S. regulations for medical devices derive from five principal laws: Federal Food, Drug and Cosmetic Act of 1938 Medical Device Amendments of 1976 Safe Medical Devices Act of 1990 Medical Device Amendments of 1992 FDA Modernization Act of 1997 (Section 204). The U.S. federal regulations that govern the testing, manufacture, and sale of medical devices are covered in Chapter 1, Title 21 of the Code of Federal Regulations (21 CFR). These comprise nine 6 × 8 inch volumes which stack 8 inches high. This title also covers foods, veterinary products, medical devices, and cosmetics. As these topics will be discussed elsewhere in this book, here we will briefly review those parts of 21 CFR that are applicable to medical devices (Gad, 2001; Heller, 1999). Of most interest to a toxicologist working in this arena would be Chapter 1, Subchapter A (Parts 1–78), which cover general provisions, organization, etc. The good laboratory practices (GLPS) are codified in 21 CFR 58. The regulations applicable to medical devices are covered in Subchapter H, Parts 800–895 of 21 CFR. As discussed earlier, the term medical device covers a wide variety of products: contact lenses, hearing aids, intrauterine contraceptive devices, syringes, catheters, drip bags, orthopedic prostheses, etc. The current structure of the law was established by the Medical Device Amendment of 1976. Products on the market 7

8

Safety Evaluation of Medical Devices

The expansion and increased sophistication of ISO guidances has tended to shift the balance towards an increasing set of required pre-IDE biocompatibility tests. In order to obtain a license to market a device, a sponsor either submits a 510(k) premarket notification or applies for a Premarket Approval (PMA), as described in 21 CFR 814. Like an NDA, a PMA application is a very extensive and detailed document that must include, among other things, a summary of clinical laboratory studies submitted in the application 921 CFR 814.20(b)(3)(v)(A), as well as a section containing results of the nonclinical laboratory studies with the device, including microbiological, toxicological, immunological, biocompatibility, stress, wear, shelf life, and other laboratory or animal tests as appropriate. As with drugs, these tests must be conducted in compliance with the GLP Regulations. Under the language of the law, a sponsor submits a PMA, which the FDA then “files.” The filing of an application means that “FDA has made a threshold determination that the application is sufficiently complete to permit substantive review.” Reasons for refusal to file are listed in 814.44(e), and include items such as an application that is not complete and has insufficient justification for the omission(s) present. The agency has 45 days from receipt of an application to notify the sponsor as to whether or not the application has been filed. The FDA has 180 days after filing of a complete PMA (21 CFR 814,40) to send the applicant an approval order, an “approved” letter or a “not approved” letter, or an order denying approval. An “approval order” is self-explanatory and is issued if the agency finds no reason (as listed in 814.45) for denying approval. An “approved” letter 814.44(e) means the application substantially meets requirements, but some specific additional information is needed. A “not approved” letter, 814.45(f), means that the application contains false statements of fact, does not comply with labeling guidelines, or that nonclinical laboratory studies were not conducted according to GLPs, etc. Essentially, an order denying approval means that the sponsor must do substantially more work and must submit a new application for PMA for the device in question. 510(k) premarket approval submissions are less extensive than PMAs, but must still include appropriate preclinical safety data. 510(k)s are supposed to be approved in 90 days. There is a third, little used route to move a new device to legal marketing approval in the U.S. This is the 513(f), filed for devices for which a 510(k) has been refused due to lack of a suitable predicate, but for which a determination of “no significant risk has been made.” Actual review and approval times historically have been much longer than the statutory limits. For 1995, the average total review time for Class III products in the United States cleared by 510(k) was 579 days (versus 240 or less in the EU) (The Gray Sheet, 1996a). For fiscal year 1996, overall average 510(k) review times (for an expected 5,875 filings) is projected to be 137 days (with low risk exempted devices and refusals to file not being included in the totals or average). Average PMA review times are projected to be 250 days (The Gray Sheet, 1996b). See Chapter 1 for a discussion of general regulatory considerations (such as Good Laboratory Practices) which are applicable to all safety evaluation studies.

Regulations Versus Law A note of caution must be inserted here. The law (the document passed by Congress) and the regulations (the documents

written by the regulatory authorities to enforce the laws) are separate documents. The sections in the law do not necessarily have numerical correspondence. For example, the regulations on the PMA process is described in 21 CFR 312, but the law describing the requirement for a PMA process is in Section 515 of the FDCA. Because the regulations rather than the laws themselves have a greater impact on the practice of nonclinical safety evaluation, greater emphasis is placed on regulation in this chapter. For a complete review of FDA law, the reader is referred to the monographs by Food and Drug Law Institute (FDLI) in 1995, 1996 and 1998. FDA http://www.fda.gov/ CDER http://www.fda.gov/cder/ CBER http://www.fda.gov/cber/index.html

Organizations Regulating Device Safety in the United States The agency formally charged with overseeing the safety of devices and diagnostics in the United States is the FDA. It is headed by a commissioner who reports to the Secretary of the Department of Health and Human Services (DHHS) and has a tremendous range of responsibilities. Medical devices are specifically overseen by the CDRH, headed by a director. Drugs are overseen primarily by the Center for Drug Evaluation and Research (CDER) (though some therapeutic or health care entities are considered as biologically derived and therefore regulated by the Center for Biologic Evaluation and Research, or CBER). There are also “combination products” (part drug, part device) which may be regulated by either or both CDER/CBER and CDRH, depending on what the principal mode of action (PMOA) is determined to be by the FDA (CFR, 1992), as discussed in Chapter 14.

Classification of Devices In the United States, in accordance with the 1976 Medical Device Amendment, devices are categorized as below. • Class I—General Controls (equivalent to OTC) • Class II—Performance Standards and Special Controls (distribution is licensed healthcare professional controlled) • Class III—Premarket Approval (clinical use only) • Preamendment Devices. In Europe, there is a lengthy set of rules in the EC Medical Device Directive (Council Directive, 1993) to place devices in Classes I, IIa, IIb or III. Class I is the minimum grade and Class II the maximum. This classification determines the extent of supporting data that is required to obtain marketing approval. In the United States, the FDA Center for Devices and Radiological Health recognizes three classes of medical device, and this system is based on whether the product was on the market prior to the passage of the 1976 Medical Device Amendments. If a new device is substantially equivalent to a pre-amendment device, then it will be classified the same as that device. This means that for Class I and II products, no premarket approval is necessary. Class III products need pre-marketing approval, and all new devices which are not substantially equivalent to existing products fall automatically into Class III. Japan (MHW) and Korea have a somewhat different three class system. Class I includes products that have no body contact and would not cause any damage to the human body if they failed, for example, x-ray film. These products need pre-marketing approval in terms of medical

Chapter 2:

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

device regulations, although they may need to be tested under industrial guidelines like those of the OECD. Class II products have external contact with the body, Class III have internal contact, and both need additional testing. Figure 2.1 presents the EFC scheme for device classification. Most of the regulatory interaction of a toxicologist involved in assessing the biocompatibility of devices is with the appropriate part of the CDRH, though for combination products the two centers charged with drugs or biologicals may also come into play. Within the CDRH there is a range of groups (called divisions) which focus on specific areas of use for devices (such as general and restorative devices; cardiovascular, respiratory, and neurological devices; ophthalmic devices; reproductive, abdominal, ear,

nose, and throat, and radiological devices; and clinical laboratory devices). Within each of these there are engineers, chemists, pharmacologists/toxicologists, statisticians, and clinicians. There is also at least one nongovernmental body which must review and approve various aspects of devices, setting forth significant “guidance” for the evaluation of safety of devices. This is the USP, and its responsibilities and guidelines are presented later in this chapter. The other two major regulatory organizations to be considered are the International Standards Organization (ISO), with ISO 10993 standards (ISO, various dates), and the Japanese Ministry of Health and Welfare (MHW) with its guidelines (MHW, 1995).

Device is a blood bag Y

START HERE Y Incorporates a medicinal substance N Blood bag N Contains animal tissue N

9

Substance is anticoagulant

Y

N N

Y

Y

III N/A

IIB

Contact with skin only

N

III

Y For contraception or prevention of sexually transmitted diseases

Y

Implantable or long term

N INVASIVE N For disinfecting, cleaning, rinsing, or hydrating contact lenses N For disinfecting medical devices N ACTIVE N To record an x-ray diagnostic image N For modifying the biological or chemical composition of blood, other body liquids or other liquids intended for infusion into the body

Y Dermis breached and can heal only by secondary intent N Used mechanical barrier, for compression, or for absorption of exudate

Figure 2.1

Medical device classification flowchart

IIB

(see page 89) Y

IIB Y IIA Y

(see page 90) Y

IIA

By filtration, centrifugation, or exchange of gas or heat

Y

For channeling or storing blood, body liquids, or gases for the purpose of eventual infusion, administration, or introduction into the body

For injured skin

III

N

Y

N

N

Y

N IIB Y IIA

Y

Y

May be connected to an active device in Class IIA or higher Y N

For storing or channeling blood or other body liquids or for storing organs, parts of organs, or body tissues N N Y

IIA Y IIA I I IIB

N IIA Y

I

(Continued)

10

Safety Evaluation of Medical Devices

Invasive with respect to a natural orifice N In oral cavity as far as the pharynx, in an ear canal up to the ear drum, or in a nasal cavity and not liable to be absorbed by Y the mucous membrane

Y

Y Long-term N

IIA

Y

Connected to an active medical device in Class IIA or higher

N

IIB

IIA

N Y In oral cavity as far as the pharynx, in an ear canal up to the drum, or in a nasal cavity

Short-term

N

IIA

Y

I

N I Specifically to diagnose, monitor, or correct a defect of the heart, or of the central circulatory system through direct contact with these parts of the body

Y

III

N To be used in direct contact with the central nervous system Transient

N III

Y N Has a biological effect or is wholly or mainly absorbed N

Transient

N

Y

III IIB

Supplies energy in the form of ionizing radiation

N

Undergoes chemical change

N

Y

Transient or short term

N

Placed in teeth

Y

N

Y

N

Transient Long-term

Y N

Administers medicines

Y

Long-term N Short-term

N

In teeth

IIB IIB N

N

Reusable instrument

Y

Transient

N Y

Figure 2.1

Continued

III

Y

Y

N

III IIB

Y

N In a potentially hazardous manner Implant or long-term

IIB AIMD

IIB IIA IIA

IIA I

Chapter 2:

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

11

Active

To administer and/or remove medicines, body liquids, or other substances to or from the body

Y

Y

In a potentially hazardous manner

N

N Y

To emit ionizing radiation or control, monitor, or influence devices emitting ionizing radiation

N Y

To administer or exchange energy therapeutically

Y

In a potentially hazardous manner

N N To control and monitor the performance of active therapeutical devices in Class IIB or intended to influence directly the performance of such devices

Y

Specifically intended for monitoring of vital physiological parameters, where the nature of variations is such that it could result in immediate danger to the patient, for instance, variations in cardiac performance, respiration, activity of CNS

N To allow direct diagnosis or monitoring of vital physiological processes

Y

Y

N

N Diagnosis by supplying energy which will be absorbed by the human body

N Diagnosis by imaging in vivo distribution of radiopharmaceuticals

Y

By illumination of the patient’s body in the visible spectrum

N

Y

Y N

Figure 2.1

Continued

TOXICITY TESTING: MEDICAL DEVICES In a statutory sense, any item promoted for a medical purpose which does not rely on chemical action to achieve its intended effect is a medical device (as discussed earlier). In vitro diagnostic tests are also regulated as medical devices. The regulation of devices under these definitions has had a different history than that of drugs—it has not been as strict and it has evolved at a slower rate. However, the requirements for the safety evaluation and biocompatibility evaluation of devices have rapidly been becoming more sophisticated and closer to that for new drugs. The safety concerns are, however, also somewhat different. Toxicologic safety concerns for devices (as opposed to concerns of

mechanical safety, such as disintegration of heart valves) are called biocompatibility concerns. Medical devices are classified as being in three different classes and are regulated accordingly. Class III devices are subject to the greatest degree of regulation and include devices which are implanted in the body, support life, prevent health impairment, or present an unreasonable risk of illness or injury. These are subject to premarketing approval. Class I and Class II devices are subject to lesser control, required only to comply with general controls and performance standards. There are several governing schemes for dictating what testing must be done on new Class III devices in the general case, with each developed and proposed by a

12

Safety Evaluation of Medical Devices

different regulatory organization at different times over the last few years. ISO has attempted to harmonize these requirements so that different (or duplicate) testing would not need to be performed to gain device approval in different national markets. As discussed in the last chapter of this book, there are also specialized testing requirements for some device types such as contact lenses (CDRH, 1995a,b) and tampons (CDRH, 1995c). The ISO effort has generally been successful and parallels that of ICH for drugs (though ISO is, it should be noted, an NGO and not a governmental regulatory body. Where differences exist, they are highlighted in this volume as specific requirements and designs are presented. As with drugs, all safety testing for devices must be conducted in conformity with GLPs (FDA, 1987; Fries, 1999; Gad and Taulbee, 1996). Table 2.1 presents the existing FDA CDRH requirements for device characterization and testing. The exact nature of the test protocols is based on recommendations by USP, ISO, and others. It should be noted that

FDA Device Categories and Suggested Biological Testing (FDA, 2000)

Circulating blood

Bone/tissue

Blood

A B C A B C A B C

• • • • • • • • •

• • • • • • • • •

• • 0 • 0 0 • • •

• • • 0 0 0 0 • •

A B C A B C

• • • • • •

• • • • • •

• 0 0 • • •

0 0 0 • • •

•

0 0

0

0 •

0 •

0

0

0 •

•

• 0

0 0

• •

• •

0 •

• •

0 0

• •

0 0

• •

0 •

• •

• • • • •

• • • •

Chronic toxicity

0 •

Hemocompatibility

0 0 0 0 0

Carcinogenicity

• • • • • • • • •

Sensitization • • • • • • • • •

Implantation

Tissue/bone dentin communicating

• • • • • • • • •

Genotoxicity

Blood path indirect

A B C A B C A B C

Subchronic toxicity

Implant devices

External Communicating devices

Breached comprised surface

Irritation or intracutaneous reactivity

Mucosal membrane

Cytotoxicity

Surface devices

Skin

Supplemental evaluation

Initial evaluation

Contact duration

Body contact

Device categories

Systemic toxicity (acute) pyrogonicity

Table 2.1

Class I devices, if new, are also subject to the ISO guidelines. It should also be noted that the FDA generally (but not strictly) now adheres to the ISO guidance on test requirements (Tables 2.12 and 2.13) Additional concerns with devices are considerations of their processing after production. For example, concerns have risen about the potential for allergies to develop to latex components and for male reproductive effects for DEHP leaching from medical devices have led to the requirement that all such devices in either of these categories be appropriately labeled. Devices which have systemic exposure need to be sterilized. Radiation and heat can be used for some devices, but others cannot be sterilized in these. Ethylene oxide or other chemical sterilants must be used, raising concerns that residual sterilants may present problems. At the same time, devices with exposure to the fluid path must be demonstrated to be neither pyrogenic nor hemolytic in their final manufactured form.

• •

•

•

0

•

•

•

•

•

•

•

A = Limited exposure (≤ 24 hours) B = Prolonged exposure (24 hours – 30 days) C = Permanent contact (>30 days) • = FDA and ISO evaluation tests 0 = Additional tests for FDA a For these devices with possible leachables or degradation products, e.g., absorbable surfaces, hemostatic agents, etc., testing for pharmacokinetics may be required. b Reproductive and developmental toxicity tests may be required for certain materials used for specialized indications. c Considerations should be given to long-term biological tests where indicated in the table taking into account the nature and mobility of the ingredients in the materials used to fabricate the device.

Chapter 2:

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

1. The selection of material(s) to be used in device manufacture and its toxicological evaluation should initially take into account full characterization of the material, for example, formulation, known and suspected impurities, and processing. 2. The material(s) of manufacture, the final product, and possible leachable chemicals or degradation products should be considered for their relevance to the overall toxicological evaluation of the device. 3. Tests to be utilized in the toxicological evaluation should take into account the bioavailability of the bioactive material, i.e., nature, degree, frequency, duration, and conditions of exposure of the device to the body. This principle may lead to the categorization of devices which would facilitate the selection of appropriate tests. 4. Any in vitro or in vivo experiments or tests must be conducted according to recognized good laboratory practices followed by evaluation by competent informed persons. 5. Full experimental data, complete to the extent that an independent conclusion could be made, should be available to the reviewing authority, if required. 6. Any change in chemical composition, manufacturing process, physical configuration or intended use of the device must be evaluated with respect to possible changes in toxicological effects and the need for additional toxicity testing. 7. The toxicological evaluation performed in accordance with this guidance should be considered in conjunction with other information from other nonclinical tests, clinical studies, and postmarket experiences for an overall safety assessment.

Device Categories: Definitions and Examples A. Noncontact Devices Devices that do not contact the patient’s body directly or indirectly; examples include in vitro diagnostic devices. B. External Devices 1. Intact surfaces Devices that contact intact external body surfaces only; examples include electrodes, external prostheses, and monitors of various types. 2. Breached or compromised surfaces Devices that contact breached or otherwise compromised external body surfaces; examples include ulcer, burn and granulation tissue dressings or healing devices, and occlusive patches. C. Externally Communicating Devices 1. Intact natural channels Devices communicating with intact natural channels; examples include contact lenses, urinary catheters, intravaginal and intraintestinal devices (sigmoidoscopes, colonoscopes, stomach tubes, gastroscopes), endotracheal tubes, and bronchoscopes. 2. Bloodpath, indirect Devices that contact the blood path at one point and serve as a conduit for fluid entry into the vascular system; examples include solution administration sets, extension sets, transfer sets, and blood administration sets. 3. Blood path, direct Devices that contact recirculating blood; examples include intravenous catheters, temporary pacemaker electrodes, oxygenators, extracorporeal oxygenator tubing and accessories, and dialyzers, dialysis tubing and accessories. D. Internal Devices 1. Bone Devices principally contacting bone; examples include orthopedic pins, plates, replacement joints, bone prostheses and cements.

13

2. Tissue and tissue fluid Devices principally contacting tissue and tissue fluid or mucus membranes where contact is prolonged; examples include pacemakers, drug supply devices, neuromuscular sensors and stimulators, replacement tendons, breast implants, cerebrospinal fluid drains, artificial larynx, vas deferens valves, ligation clips, tubal occlusion devices for female sterilization, and intrauterine devices. 3. Blood Devices principally contacting blood; examples include permanent pacemaker electrodes, artificial arteriovenous fistulae, heart valves, vascular grafts, blood monitors, internal drug delivery catheters, and ventricular assist pumps.

Biological Tests Also required to properly utilize the tables is a knowledge of the objectives of the specified biological tests. These can be considered as follows (Gad and Chengelis, 1998; Goering and Galloway, 1989): Sensitization Assay Estimates the potential for sensitization of a test material and/or the extracts of a material using it in an animal and/or human. ISO (ISO, 1992, 1996) and MHW procedures are contrasted in Table 2.2. Irritation Tests Estimates the irritation potential of test materials and their extracts, using appropriate site or implant tissue such as skin and mucous membrane in an animal model and/or human. ISO and MHW procedures are contrasted in Table 2.3; and for eye irritation in Table 2.4. Cytotoxicity With the use of cell culture techniques, this test determines the lysis of cells (cell death), the inhibition of cell growth, and other toxic effects on cells caused by test materials and/or extracts from the materials. ISO and MHW procedures are contrasted in Table 2.5. Acute Systemic Toxicity Estimates the harmful effects of either single or multiple exposures to test materials and/or extracts, in an animal model, during a period of less than 24 hours. ISO and MHW procedures are contrasted in Table 2.6. Hematocompatibility Evaluates any effects of blood contacting materials on hemolysis, thrombosis, plasmaproteins, enzymes, and the formed elements using

Table 2.2 Differences Between Sensitization Test Procedures Required by ISO 10993-10 and the MHW Guidelines ISO 10993-10

MHW 1995

Sample preparation: Extraction in polar and/or nonpolar solvents.

Two extraction solvents, methanol and acetone, recommended.

Extraction ratio: Extraction ratio is dependent on thickness of device or representative portion. Extract used for testing. If extraction is not possible, the adjuvant and patch test can be utilized.

Specific extraction ratios: 10:1 (volume solvent:weight sample)

Residue obtained from extraction is redissolved and used for testing. (If residue does not dissolve in DMSO, or a sufficient amount of residue is not obtained, the adjuvant and patch test is recommended). Sufficient amount of residue: 0.1–0.5% (weight residue:weight test material)

14

Safety Evaluation of Medical Devices

Table 2.3 Differences in Intracutaneous Reactivity Test Procedures Required by ISO 10993-10 and the MHW Guidelines

Table 2.5 Differences Between Cytotoxicity Test Procedures Specified by ISO 10993-5 and the MHW Guidelines (MHW, 1995)

ISO 10993-10

MHW

ISO 10993-10

MHW 1995

Two rabbits for each extract.

Number of cells per dish: 0.5–1 million cells

40 to 200 cells per dish

Number of test animals: Three rabbits for 1 to 2 extracts. Number of test/control injections per extract: Five test and five control injections.

Ten test and five control injections.

Evaluation of responses: Quantitative comparison of responses of test and control responses.

Qualitative comparison of test and control responses.

an animal model. Traditionally, hemolysis, which determines the degree of red blood cell lysis and the separation of hemoglobin caused by test materials and/or extracts from the materials in vitro, has been “the” representative test employed. A broader range of primary tests (adding evaluations of thrombosis, coagulation, platelets, and immunology aspects) is currently recommended. ISO and MHW procedures for hemolysis are contrasted in Table 2.7. Implantation Tests Evaluates the local toxic effects on living tissue, at both the gross level and microscopic level, to a sample material that is surgically implanted into appropriate animal implant site or tissue, e.g., muscle, bone; for 7–90 days. ISO and MHW procedures are contrasted in Table 2.10. Mutagenicity (Genotoxicity) The application of mammalian or non-mammalian cell culture techniques for the determination of gene mutations, changes in chromosome structure and number, and other DNA or gene toxicities caused by test materials and/or extracts from materials. Selected tests representing gene mutation tests (Ames or mouse lymphoma), chromosomal aberration tests (CHO) and DNA effects tests (mouse micronucleous and sister chromatid exchange) should generally be employed. ISO and MHW procedures are contrasted in Table 2.10. Subchronic Toxicity The determination of harmful effects from multiple exposures to test materials and/or extracts during a period of one day to less than 10% of the total life of the test animal (e.g., up to 90 days in rats). Chronic Toxicity The determination of harmful effects from multiple exposures to test materials and/or extracts during a period of 10% to the total life of the test animal (e.g., over 90 days in rats). Carcinogenesis Bioassay The determination of the tumorigenic potential of test materials and/or extracts from either single or multiple exposures, over a period of the total life (e.g., 2 years for rat, 18 months for mouse, or 7 years for dog). Pharmacokinetics To determine the metabolic processes of absorption, distribution, biotransformation, and Table 2.4 Differences in Eye Irritation Testing Procedures Outlined in ISO 10993-10 and the MHW Guidelines ISO 10993-10

MHW 1995

Time of exposure: 1 second

Thirty seconds

Grading scale: Classification system for grading ocular lesions

Draize or McDonald–Shadduck scale.

Extraction ratio: 60 cm2 per 20 mL if thickness 80.5 mm 120 cm2 per 20 mL if thickness 70.5 mm or 4g per 20 mL

5 cm2/mL or 1 g/10 mL

Exposure period: Typically 24–72 hours (2 hours for filter diffusion test)

6–7 days

Toxicity determination: Visual grading and/or quantitative assessments

Quantification of surviving colonies

Positive controls: Materials providing a reproducible cytotoxic response (e.g., organo-tin-impregnated polyvinyl chloride)

Segmented polyurethane films containing 0.1% zinc diethyldithiocarbamate and 0.25% zinc dibutyldithiocarbamate

elimination of toxic leachables and degradation products of test materials and/or extracts. Reproductive and Developmental Toxicity The evaluation of the potential effects of test materials and/or extracts on fertility, reproductive function, and prenatal and early postnatal development. The tests for leachables such as contaminants, additives, monomers, and degradation products must be conducted by choosing appropriate solvent systems that will yield a maximal extraction of leachable materials to conduct biocompatibility testing. Chapter 3 addresses the issues behind sampling, sample preparation, and solvents. The effects of sterilization on device materials and potential leachables, as well as toxic by-products, as a consequence of sterilization should be considered. Therefore, testing should be performed on the final sterilized product or representative samples of the final sterilized product. Table 2.10 presents the basis for test selection under the Tripartite Agreement.

United States Pharmacopoeial Testing The earliest guidance on what testing was to be done on medical devices was that provided in the USP and other pharmacopoeias. Each of the major national pharmacopoeias offers somewhat different guidance. The test selection system for the USP (presented in Table 2.10), which classified plastics as Classes I through VI, is now obsolete and replaced in usage by the other guidelines presented here. But the actual descriptions of test types, as provided in the USP (and presented in the appropriate chapters later in this book) are still very much operative (USP, 1994). There are British, European, and Japanese pharmacopoeias, of which the latter requires the most attention due to some special requirements still being operative if product approval is desired.

ISO Testing Requirements The European Economic Community has adopted a new set of testing guidelines for medical devices under the aegis of

Chapter 2:

Table 2.6

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

15

Comparison of Grading Scales Used to Score Responses of Test Animals to ASTM and ISO/USP Procedures

Response Normal, no symptoms Slight Moderate

Marked

Dead, expired

ASTM Description

ISO/USP

Mouse exhibits no adverse physical symptoms after injection. Mouse exhibits slight but noticeable symptoms of hypokinesis, dyspnea, or abdominal irritation after injection. Mouse exhibits definite evidence of abdominal irritation, dyspnea, hypokinesis, ptosis, or diarrhea after injection. (Weight usually drops to between 15 and 17 g.) Mouse exhibits prostration, cyanosis, tremors, or severe symptoms of abdominal irritation, diarrhea, ptosis, or dyspnea after injection. (Extreme weight loss; weight usually less than 15 g.) Mouse dies after injection. Interpretation The test is considered negative if none of the animals injected with the test article extracts shows a significantly greater biological reaction than the animals treated with the control article.

Interpretation The test is considered negative if none of the animals injected with the test article shows a significantly greater biological reaction than the animals treated with the control article.

If two or more mice show either marked signs of toxicity or die, the test article does not meet the requirements of the test.

If two or more mice die, or show signs of toxicity such as convulsions or prostration, or if three or more mice lose more than 2 g of body weight, the test article does not meet the requirements of the test.

If any animal treated with a test article shows slight signs of toxicity, and not more than one animal shows marked signs of toxicity or dies, a repeat test using freshly prepared extract should be conducted using groups of 10 mice each. A substantial decrease in body weight for all animals in the group, even without other symptoms of toxicity, requires a retest using groups of 10 mice each. In the repeat test, the requirements are met if none of the animals injected with the test article shows a substantially greater reaction than that observed in the animals treated with the control article.

If any animal treated with a test article shows only slight signs of biological reaction, and not more than one animal shows gross signs of biological reaction or dies, a repeat test should be conducted using groups of 10 mice. On the repeat test, all 10 animals must not show a significantly greater biological reaction than the animals treated with the control article.

ISO (ISO, 1992; The Gray Sheet, 1992). The ISO 10993 guidelines for testing provide a unified basis for international medical device biocompatibility evaluation, both in terms of test selection (as presented in Tables 2.11 and 2.12) and test design and interpretation (Table 2.13). In 1996, the United States FDA also announced that it would adhere to ISO 10993 standards for device biocompatibility evaluation. This international standard specifies methods of biological testing of medical and dental materials and devices and their evaluation in regard to their biocompatibility. Because of the many materials and devices used in these areas, the standard offers a guide for biological testing.

Actual test performance standards also vary, as shown in Tables 2.3–2.10. Committees dealing with materials and devices must decide on tests and test series relevant to the respective materials and devices, It is the responsibility of the product committees to select adequate test methods for products. The standard contains animal tests, but tries to reduce those tests to the justifiable minimum. Relevant international and national regulations must be observed when animals are used.

MHW Requirements

Table 2.8 Comparison of Pyrogen Test Procedures Required by ISO 10993-11 and the MHW Guidelines

The Japanese ISO test selection guidelines vary from those of FDA and ISO and are summarized in Table 2.15 (MHW, 1995; Japanese Pharmacopoeia, 1996).

Table 2.7 Differences in Hemolysis Test Procedures Recommended by ISO 10993-4 and the MHW Guidelines ISO 10993-4

MHW 1995

Hemolysis can be assessed by any of several validated methods to assay hemoglobin in plasma.

Hemolytic index is assessed by measuring hemoglobin at 1, 2, and 4 hours by spectrophotometric methods. The hemolysis over this period is expressed as a percentage of the positive control.

ISO 10993-11

MHW 1995

Number of animals: Three rabbits required; comparison of febrile response in test animals to baseline temperature for evaluation of pyrogenicity potential

Three rabbits (test) required; comparison to baseline temperature is evaluated as index of pyrogenicity potential

Test duration: Test measurement intervals: every 30 minutes for 3 hours

Test measurement intervals: every hour for 3 hours

Evaluation: Cutoff for positive febrile response: 0.5°C

Cutoff for positive febrile response: 0.6°C

16

Safety Evaluation of Medical Devices

Table 2.9 Differences in ISO 10993-3 and the MHW Guidelines for Assessing the Effects of Device or Material Implantation ISO 10993-3

MHW 1995

Time point(s) of assessment: Sufficient to achieve steady state (e.g. 2, 4, 6, and 12 weeks)

7 days and 4 weeks

Number of animals: At least three per time period of assessment

At least four per time period

Number of samples of evaluation: At least eight per time period for test and control

No minimum number specified

Evaluation criteria: Comparative evaluation of responses to test and control materials

If more than two of the four test sites in each animal exhibit a significant response compared to control sites, the test is considered positive

ISO 10993 is based on existing national and international specifications, regulations, and standards wherever possible. It is open to regular review whenever new research work is presented to improve the state of scientific knowledge. Tables 2.3 and 2.4 provide the test matrices under ISO 10993. Subsequently, specific guidance on individual test designs, conduct and interpretation has been provided as subparts 2-11 of ISO-10993 (Table 2.13) (AAMI, 2006).

CE Marking of Devices After June 14, 1998, all medical products distributed in Europe have had to bear the CE mark. ISO 9000 certification supplements and supports an assessment of conformity to the Medical Devices Directive (MDD), which must be performed by a certification body appointed by the EU member states (Haindl, 1997). To qualify for the CE mark, manufacturers of Class IIa, IIb, and III devices must be certified by a notified body (which is recognized by the national health authorities) to Annex II, V, or VI of the MDD (also known as 93/42/EEC) and comply with the essential requirements of the directive. Manufacturers of active implantables and IVDs have separate directives to contend with. When auditing for compliance, the notified body will check a number of items in addition to a manufacturer’s QA system, including technical files, sterility assurance measures, subcontracting procedures, recall and vigilance systems, and declarations of

Table 2.10

conformity. Depending on the classification and certification route, some devices will also require an EC-type examination or a design review by the notified body. Manufacturers of Class I products, who require minimal interaction with a notified body, appear to be the clear winners in this scheme, but even they must deal with a number of vague or confusing requirements (Table 2.15). Simply classifying their products according to the dictates of 93/42/EEC, Annex IX, can be a tricky affair, and faulty classification can lead to bigger problems. The simplified flowcharts in Figure 2.1 should help manufacturers determine whether their products qualify as Class I devices. For more difficult products, manufacturers may need to refer to a consultant or obtain a suitable software program. Classification is based on the intended and declared use of a product, not solely on its salient features. The Class I designation usually—but not always—excludes sterile products and measuring devices that measure physiological parameters or require a high degree of accuracy. So, for example, a reusable scalpel is Class I, but a sterile scalpel is Class IIa; a scalpel blade for the reusable device is Class I, but if it is supplied sterile, it is Class IIa; a scalpel blade for the reusable device is Class I but if it is supplied sterile, it is Class IIa. A stethoscope, a simple graduated syringe (not for injection pumps), and a measuring spoon for administering an expectorant are not considered measuring devices, although a hand-driven blood-pressure gage and a digital thermometer are. All of the classification rules are included in the directive, but they are not easy to understand. An EC working group has drawn up a separate paper known as MEDDEV 10/93 to explain the rules and provide some practical guidelines. For example, the directive stipulates that reusable surgical instruments belong in the Class I designation as long as they are not intended for more than an hour of continuous use. According to this definition, items such as scissors and tweezers, even if they are used in a six-hour operation, are still considered Class I devices because they are not used continuously during that time. Even if a Class I product is supplied sterile, the manufacturer must issue a self-declaration of conformity. In this case, the manufacturer need only certify the QC system governing those aspects of manufacture concerned with securing and maintaining sterile conditions. If the device is packaged and sterilized by a company that works with a certified process, then the manufacturer must only validate the process for the particular device and submit the results to

Differences in Genotoxicity Testing Procedures Required by ISO 10993-3 and the MHW Guidelines

ISO 10993-10

MHW 1995

Extraction vehicles: A physiological medium is used and, where appropriate, a solvent (e.g., dimethylsulfoxide)

Recommends methanol and acetone as extracting vehicles

Extraction: Extract test material and test the extract or dissolve material in solvent and conduct test. The conditions of extraction should maximize the amount of extractable substances, as well as subject the test device or material to the extreme conditions it may be exposed to, without causing significant degradation. Extraction ratio is dependent on thickness of test material.

Extract at room temperature at a ration of 10:1 (solvent:material) and obtain residue (at least 0.1–0.5% [weight of residue/weight of test material]), redissolve in appropriate solvent and test residue. If sufficient residue is unobtainable, extract test material (in ethanol, acetone, or DMSO at 10 g of test material per 20 mL for the Arnes mutagenicity assay, and in cell culture medium at 120 cm3 or 4 g/20 mL for the chromosomal aberration assay), at 37°C for 48 hours and test extract. The Ames mutagenicity assay is conducted with a volume of 200 µL per plate.

x x x x x

x x x x

x

x x

x x x x

III x x x x x x x x

IV

Plastic classesa

x x

x x x x x

V

Classification of Plastics (USP XXIII)

x x x

x x

VI

Implant strips of sample

Extract of sample in vegetable oil

Extract of sample in 1 in 20 Solution of alcohol in sodium chloride injection Extract of sample in polyethylene glycol 400

Extract of sample in sodium chloride inspection

Test material Mouse Rabbit Mouse Rabbit Mouse Rabbit Mouse Rabbit Rabbit

Animal

50 mL/kg 0.2 mL/animal at each of 10 sites 50 mL/kg 0.2 mL/animal at each of 10 sites 10 g/kg 0.2 mL/animal at each of 10 sites 50 mL/kg 0.2 mL/animal at each of 10 sites 4 strips/animal

Dose

Tests to be conducted

A (ip) B C

A (ip)

A (iv) B A (iv)

Proceduresb

b

Tests required for each class are indicated by “x” in appropriate rows. Legend: A (ip), Systemic injection test (intraperitoneal); A(iv), Systemic injection test (intravenous); B, Intracutaneous (Intracutaneous); C, Implantation test (intramuscular implantation). The table lists the biological tests that might be applied in evaluating the safety of medical devices and/or polymers. This does not imply that all the tests listed under each category will be necessary or relevant in all cases. Tests for devices made of metals, ceramics, biological materials, etc., are not included here but are under consideration. Categorization of medical devices is based on body contact and contact duration.

a

II

I

Table 2.11

Safety Evaluation of Medical Devices

ISO Initial Evaluation Tests

Irritation or intracutaneous

x x x x x x x x x

x x x x x x x x x

x x x x x x x x x

A B C A B C

x x x x x x

x x x x x x

x x

A B C

x x x

x x x

x x x

A B C A B C

x x x x x x

x x x x x x

x

x

x

x

x

Hemocompatability

Sensitization

A B C A B C A B C

Implantation

Cytotoxicity

Body contact duration A—limited exposure B—prolonged or repeated exposure C—permanent contact

Pyrogenicity

Biological tests

Acute systemic toxicity

Device categories

Mutagenicity

Table 2.12

Subchronic toxicity

18

Surface devices: Skin

Mucous membranes

Breached surface Externally communicating: Blood path indirect

Tissue/bone communicating

x

x

x

x x x

x x x

x x x

x x

Internal devices: Circulating blood

x x x

x

x x

x x x

x x x

Implant devices: Bone/tissue

Blood

a notified body. The manufacturer still needs certification by a notified body in regard to the performance aspects relating to sterility and measurement function; the notified body will also want to inspect the manufacturer’s facility. Nonetheless, the procedure is far less complicated than a full production audit. All manufacturers applying for CE marking privileges— including manufacturers of Class I devices—must prepare the proper technical documentation; appoint a “responsible person” within the EEC; design product labels and labeling according to 93/42/EEC, Annex I, paragraph 13; and sign a declaration of conformity. The technical dossier should not pose a major problem for manufacturers familiar with device master files. A list of required dossier contents is given in Table 2.16. For biological material testing, Europe uses the ISO 10993 (EN 30993) protocols, but test results according to the Tripartite agreement (or USP XXIII) are accepted. Every electrical device must also be proven to comply with the EMC requirements defined in the MDD; suppliers of preassembled electrical components may have the appropriate test results already available. Reformatting an existing device master file is not necessary, only creating an index

x x x x x

x x x

x

x x

x x x

x x x x x

x x x

that cross-references the essential requirements of the directives with the device file contents. The master file is a controlled document, as defined in ISO 9000, and manufacturers would do well to regard it as highly confidential. The technical dossier is closely linked to the responsible person, a representative in the EEC governed by European law and authorized by the manufacturer to oversee routine regulatory affairs. Specifically, the responsible person must ensure compliance with the European vigilance system, which covers both postmarket surveillance and adverseincident reporting. For example, if a patient were injured by a device, or if a patient would have been injured had the caregiver not intervened, the responsible person would have to investigate the incident together with the device’s manufacturer and file a report with the competent authorities. Moreover, the European authorities must be able to obtain the master file in case of trouble; therefore, the manufacturer must either store the file or its abbreviated form with the responsible person or draw up a contractual agreement that gives the agent the right to access the master file without delay if required by the authorities. The agent must be available all year, as the time frame for notification could be as

Chapter 2:

19

ISO Special Evaluation Tests Biological tests

Chronic toxicity

Carcinogenicity

Body contact duration A—limited exposure B—prolonged or repeated exposure C—permanent contact (time limits to added)

x

x

Degradation

Device categories

Reproductive/developmental

Table 2.13

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

Surface devices: Skin

Mucous membranes

Breached surface

A B C A B C A B C

Externally communicating: Blood path indirect

Tissue/bone communicating

A B C A B C

x

Internal devices: Circulating blood

Bone/tissue

Blood

Table 2.14

A B C A B C A B C

x

x

x

x

x

X

ANSI/AAMI/ISO Standards

Evaluation and testing Animal welfare requirements Tests for genotoxicity, carcinogenicity and reproductive toxicity Selection of tests for interactions with blood Tests for in vitro cytotoxicity Tests for local effects after implantation Tests for irritation and delayed-type hypersensitivity Ethylene oxide sterilization residuals CANCELLED Framework for identification and quantification of potential degradation products Tests for systemic toxicity Sample preparation and reference materials Identification and quantification of degradation products from polymeric devices Identification and quantification of degradation products from ceramics Identification and quantification of degradation products from metals and alloys Toxicokinetic study design for degradation products and leachables from medical devices Establishment of allowable limits for leachable substances Physio-chemica, morphological and topographical characterization of materials Chemical characterization of materials Principles and methods for immunotoxicology testing of medical devices Clinical Investigation of medical devices for human subjects-Part 1: General requirements Clinical Investigation of medical devices for human subjects—Part 2: Clinical Investigation plans

ISO designations

Year issued

10993-1 10993-2 10993-3 10993-4 10993-5 10993-6 BE78 10993-7 10993-8 10993-9 10993-11 10993-12 10993-13 10993-14 10993-15 10993-16 10993-17 10993-19 BE83 10993-20 14155-1 14155-2

2003 2006 2003 2002 and A1/2006 1999 1995/(R)2001 2002 1995/(R)2001 — 1999/(R)2005 2006 2002 1999/(R)2004 2001 2000 1997/(R)2003 2002 2006 2006 2006 2003 2003

20

Safety Evaluation of Medical Devices

Japanese MHW Test Selection Guidelines

Tissue/bone dentin communicating

Circulating blood

Bone/tissue

Blood

A = Temporary contact (<24 hours)

X X X X X X X X X

X X

A B C A B C

X X X X X X

X X X X X X

X

X

X

X

X

X X X

X X X

X

X

X X X X X X

X X X

X

X X

X X X X X

X X X X X

X X X

X

X X

B = Short- and medium-term contact (24 hours–29 days)

short as 10 days. Ideally, the responsible person should be familiar with the national regulation in all member states. The simplest way to maintain a European address will be to appoint a distributor as their responsible person, although this course is not without potential problems. The selected distributor does not need certification as long as the manufacturer’s name and CE mark are on the product labeling. The name of the responsible person must also appear on the label, package insert, or outer packaging, even if the product is sold by a completely different distributor in another country. There is no official rule or proposal regarding Table 2.16

X X X

Carcinogenicity

X X X X X X X X X

X

Chronic toxicity

A B C A B C A B C

X

Implantation

X X X X X X X X X

Pyrogen

X X X X X X X X X

Genotoxicity

X X X X X X X X X

Subchronic toxicity

A B C A B C A B C

Hemocompatibility

Blood path indirect

Irritation or intracutaneous reactivity

Implant devices

External Communicating devices

Breached/compromised surface

Sensitization

Mucosal membrane

Cytotoxicity

Surface devices

Skin

Supplemental evaluation

Initial evaluation

Contact duration

Body contact

Device categories

Systemic toxicity (acute) pyrogonicity

Table 2.15

X X X

X X X X

X X X X X

X X X

X

X

X

X

X

X

C = Long-term contact (>30 days)

how many responsible persons a manufacturer should have, but each one must appear on the labeling; therefore, appointing more than one is of limited use. The responsible person should be selected with great care; device master files (Table 2.16) must be made available to the responsible person in the event of patient injury or near injury, and many distributors are potential competitors. Class I devices, by nature, will rarely lead to patient injury, but manufacturers should still consider labeling issues when choosing a representative. It is easy to change distributors, but changing the responsible person means changing all the product labeling.

Which Products are Class I?

The classification of a product refers to its intended use. The following is a simplified listing of Class I products: • Noninvasive (and nonactive) devices that do not modify the biological or chemical composition of blood or liquids intended for infusion; store blood, body liquids, or tissues for administration; or connect to an active medical device. • Dressings intended only as a mechanical barrier or for absorption of exudates. • Invasive products for use in natural body orifices and stomas for no longer than one hour or in the oral or nasal cavity or ear canal for up to 30 days. • Surgical invasive products if they are reusable instruments and not intended for continuous use of more than one hour. • Active devices that administer neither energy nor substances to the body nor are made for diagnosis. Class I products cannot: • Incorporate medicinal products (drugs) or animal tissue. • Be intended for contraception or the prevention of sexually transmitted diseases.

Chapter 2:

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

As an alternative, manufacturers can contract with a professional agency to serve as a representative completely independent from any distribution network. The issue of labeling is itself a source of contention. Not all countries have decided yet whether they will insist on having their own language on device labels. Many countries have rather imprecise rules, dictating that their national language must appear only if necessary. Manufacturers can reduce potential trouble by using the pictograms and symbols defined in the harmonized European standard EN 980. For instructions of use, manufacturers are advised to use all 12 languages used in the European Economic Area. The requirements for labeling are presented in Annex I, paragraph 13, of the MDD; some devices may be subject to additional requirements outlined in product standards. Class I products fall under the jurisdiction of local authorities, but who serves as those authorities may differ from country to country. In Germany, for example, there are no clear-cut regulations that define the competence of the local authorities, except in the case of danger to the patient. European product liability laws more or less give the consumer the right to sue anybody in the trade chain. Normally, claims would be filed against the manufacturer, but it is possible that there will be claims against a responsible person. This is a rather new legal situation, and the rules will be determined by court decisions. It is hoped that Class I products will not instigate many court actions, but clearly, even manufacturers of Class I devices will have a host of new concerns under the CE marking scheme. Table 2.17

Contents of a Device Master File

1. EC declaration of conformity and classification according to Annex IX of the MDD. 2. Name and address of the manufacturer’s European responsible person. 3. Product description, including: • All variants. • Intended clinical use. • Indications/contraindications. • Operating instructions/instructions for use. • Warnings/precautions. • Photographs highlighting the product. • Photographs highlighting the usage. • Brochures, advertising, catalog sheets, marketing claims (if available). • Product specifications including: • Parts list, list of components. • Specifications of materials used, including data sheets. • List of standards applied. • Details of substance(s) used (in the event of drug-device combination). • QA specifications (QC specs, in-process controls, etc) etc. • Labeling, accompanying documents, package inserts (DIN EN 289, prEN 980). • Instruction for use (prEN 1041). • Service manual. • Product verification, including: • Testing data and reports, functionality studies, wet lab or benchtop testing. • Materials certificates/reports on biological tests. • EMC testing and certificates. • Validation of the packaging/aging studies. • Compatibility studies (connection to other devices). • Risk analysis (DIN EN 1441). • Clinical experience. 4. List of requirements (Annex I) indicating cross-reference with documentation.

21

Risk Assessment The reality is that not all materials used on devices are entirely safe. Generally, if one looks long enough at small enough quantities, some type of risk can be associated with every material. Risk can be defined as the possibility of harm or loss. Health risk, of course, is the possibility of an adverse effect on one’s health. Risk is sometimes quantified by multiplying the severity of an event times the probability the event will occur, so that: Risk = severity × probability While this equation appears useful in theory, in practice it is difficult to apply to the biological safety of medical devices. The process known as health-based risk assessment attempts to provide an alternative strategy for placing health risks in perspective (Stark, 1998; AAMI, 1998).

Standards and Guidances A paradigm for the risk assessment process has been detailed in a publication prepared by the U.S. National Academy of Sciences (Hayes, 1994). Although devised primarily for cancer risk assessment, many of the provisions also apply to the assessment of other health effects. The major components of the paradigm are (1) hazard identification, (2) dosage-response assessment, (3) exposure assessment, and (4) risk characterization (Ecobichon, 1992). The general approach to risk assessment was adapted to medical devices via the draft CEN standard Risk Analysis, published in 1993,a and more recently via the ISO standard, ISO 14538—Method for the establishment of Allowable Limits for Residues in Medical Devices Using Health-Based Risk Assessment, published in 1996.b At the present time, the FDA is also working to develop a health-based risk assessment protocol adapted to medical devices. Informally called the Medical Device Paradigm, the document is not yet generally available (Brown and Stratmeyer, 1997).c Some manufacturers may object that regulators are once again attempting to impose a “drug model” on medical devices. However, we shall see in the following pages that judicious application of these risk assessment principles can provide a justification for using materials that carry with them some element of risk, and that may, under traditional biocompatibility testing regimes, be difficult to evaluate or be deemed unsuitable for medical device applications.

Method Hazard Identification The first step in the risk assessment process is to identify the possible hazards that may be presented by a material. This is accomplished by determining whether a compound, an extract of the material, or the material itself produces adverse effects, and by identifying the nature of

a

CEN BTS 3/WG 1—Risk Analysis is available through the British Standards Institute. b Available from the Association for the Advancement of Medical Instrumentation, 3330 Washington Blvd., Ste. 400, Arlington, VA 22201, USA. c Draft copies of the Medical Device Paradigm may be obtained by contacting Dr. Melvin Stratmeyer, FDA Center for Devices and Radiological Health, HFZ-112, Division of Life Sciences, Office of Science and Technology, FDA, Rockville, MD 20857, USA.

22

Safety Evaluation of Medical Devices

those effects. Adverse effects are identified either through a review of the literature or through actual biological safety testing.

Dose-Response Assessment The second step is to determine the dose response of the material—that is, what is the highest weight or concentration of the material that will not cause an effect? This upper limit is called the allowable limit. There are numerous sources in the literature of data from which to determine allowable limits; some will be more applicable than others, and some may require correction factors.

Exposure Assessment The third step is to determine the exposure assessment by quantifying the available dose of the chemical residues that will be received by the patient. This is readily done by estimating the number of devices to which a patient is likely to be exposed in a sequential period of use (for instance, during a hospital stay) or over a lifetime. For example, a patient might be exposed to 100 skin staples following a surgical procedure, or to two heart valves in a lifetime; thus, the amount of residue available on 100 skin staples or two heart valves would be determined.

Risk Characterization Characterizing the risk constitutes the final step of the process. The allowable limit is compared with the estimated exposure: if the allowable limit is greater than the estimated exposure by a comfortable safety margin, the likelihood of an adverse event occurring in an exposed population is small, and the material may be used.

whether or not in vivo corrosion of nitinol releases unsafe levels of nickel. Dose-Response Assessment. A search of the world medical literature revealed that the recommended safe level of exposure to nickel in intravenous fluids is a maximum of 35 µg/day (Stark, 1996). This value can be taken as an allowable limit of nickel exposure for a 70-kg (154-lb) adult. The intravenous fluid data are based on subjects that are comparable to the patients who will be receiving nitinol implants. The data are for humans (not animals), for ill patients (not healthy workers or volunteers), and for similar routes of exposure (intravenous fluid and tissue contact). For these reasons, no safety correction factor need be applied to the allowable limit of exposure. Exposure Assessment. The available dose of nickel from nitinol implants can be estimated from data found in the literature. In one study, dental arch wires of nitinol were extracted in artificial saliva, and the concentration of nickel measured in the supernatant. Corrosion reached a peak at day 7, then declined steadily thereafter. The average rate of corrosion under these conditions was 12.8 µg/day/cm2 over the first 28 days. Risk Characterization. A comparison of the available dose with the allowable limit for intravenous fluid levels shows that there is approximately a threefold safety margin, assuming that the implanted device is a full 1 cm2 in surface area. (Devices with less surface area will contribute even less to the nickel concentration and have an even larger safety margin.) Considering the high quality of the data, a threefold safety margin is sufficient to justify using nitinol in vascular implants.

Wound-Dressing Formulation Case Studies We can best get a sense of how these standards work by looking at some actual medical case studies that illustrate the risk assessment process (Stark, 1997).

Nitinol Implant Nitinol is an unusual alloy of nickel and titanium that features the useful property of “shape memory.” A nitinol part can be given a particular shape at a high temperature, then cooled to a low temperature and compressed into some other shape; the compressed part will subsequently deploy to its original shape at a predetermined transition temperature. This feature is particularly beneficial for vascular implant applications in which the shape of the device in its compressed state eases the insertion process. The nitinol deploys as it is warmed by the surrounding tissue., expanding to take on the desired shape of a stent, filter, or other device. The transition temperature depends on the alloy’s relative concentrations of nickel and titanium: a typical nickel concentration of 55–60% is used in medical devices, since this gives a transition temperature at approximately the temperature of the body (37°C). Hazard Identification. One concern with using nitinol in implant applications is the potential release of nickel into the body. Although nickel is a dietary requirement, it is also highly toxic—known to cause dermatitis, cancer subsequent to inhalation, and acute pneumonitis from inhalation of nickel carbonyl, and to exert a toxic effect on cellular reproduction. It is a known sensitizer, with approximately 5% of the domestic population allergic to this common metal, probably through exposure from costume jewelry and clothing snaps. The biocompatibility question at hand is

Today’s wound dressings are highly engineered products, designed to maintain the moisture content and osmatic balance of the wound bed so as to promote optimum conditions for wound healing. Complex constructions of hydrocolloids and superabsorbers, these dressings are sometimes used in direct tissue contact over full-thickness wounds that penetrate the skin layers. Hazard Identification. There have been reports in the literature of patients succumbing to cardiac arrest from potassium overload, with the wound dressing as one of the important contributors of excess potassium in the bloodstream. The effects of potassium on cardiac function are well characterized. Normal serum levels for potassium are 3.8–4 milliequivalents per liter (mEq/L). As the potassium concentration rises to 5–7 mEq/L, a patient can undergo cardiac arrest and die. The biocompatibility issue to be explored is whether or not a wound-dressing formulation might release dangerous levels of potassium if used on full-thickness wounds. Dose-Response Assessment. An increase of approximately 1 mEq/L of potassium is unlikely to provoke mild adverse events in most patients. Assuming that the average person’s blood volume is 5 L, a one-time dose of 5 mEq of potassium may begin to cause adverse reactions. This value can be considered to be the allowable limit of potassium for most patients. Exposure Assessment. Let us suppose that each dressing contains 2.5 g of potassium bicarbonate. Since the molecular weight of potassium bicarbonate is 100 g/mole, each dressing contains 0.025 mole of sodium bicarbonate, or 0.025 mEq of potassium ion. If a patient were to use four dressings in a day, the available dose of potassium would be 0.1 mEq/day.

Chapter 2:

Regulatory Aspects and Strategy in Medical Device and Bio Materials Safety Evaluation

Risk Characterization. Comparing the available dose of potassium (0.1 mEq) to the allowable limit (5 mEq) shows that there is a 50-fold safety margin. Considering that patients may be small in size, may have kidney impairment, or may receive potassium from additional sources such as intravenous fluids, this safety margin is too small, and so the dressing should be reformulated.

Perchloroethylene Solvent A manufacturer of metal fabricated parts uses perchloroethylene to clean the finished pieces. Perchloroethylene has many advantages as a cleaner and degreaser: it is highly volatile, does not damage the ozone layer, and is very effective as a precision cleaning solvent. The most common use of perchloroethylene is in the dry cleaning industry, but it is also commonly used in the electronics industry to clean circuit boards. Hazard Identification. The downside of perchloroethylene is that is it highly toxic, with a material safety data sheet several pages in length listing adverse effects ranging from dizziness to death. Biocompatibility testing on solventcleaned parts would be meaningless; the solvent concentration on the part is so small that any effects of the solvent would be masked by the natural biological process of the test animals. The biocompatibility question that must be answered is whether or not sufficient residual perchloroethylene remains on the cleaned metal parts to pose a health hazard. Dose-Response Assessment. Threshold limit values (TLVs) are values that indicate the maximum level of a chemical that a healthy worker could take in on a daily basis over the course of his or her work life without experiencing any adverse effects (ACGIH, 1986; AHHA, 1980). The TLV for perchloroethylene is 50 ppm/day (50 mL of perchloroethylene per 103 liter of air) by inhalation. The average person inhales 12,960 L of air per day, making this equivalent to 650 mL of perchloroethylene per day. Since the vapor density of perchloroethylene is 5.76 g/L, the TLV is equal to 3.7 g of perchloroethylene per day by inhalation. Because TLVs for inhalation—as opposed to direct tissue exposure—are determined based on healthy individuals (not ill patients), we will divide the TLV by an uncertainly factor of 100, i.e., 10 to account for a different route of exposure and 10 to account for healthy-to-ill persons. By this method, we obtain an allowable perchloroethylene limit of 37 mg/day. Exposure Assessment. To calculate an available dose of perchloroethylene, we need some additional information. In this case, the manufacturer brought a number of cleaned metal pieces into equilibrium within a closed jar, then analyzed the headspace above the pieces by using a highpressure liquid chromatography to determine the concentration of perchloroethylene released. The concentration of perchloroethylene was undetectable by high-performance liquid chromatography. Since the limits of this analytical method are 2 ppb, this value was taken as the concentration of perchloroethylene in the headspace. Taking the weight of the metal pieces, the number of pieces tested, and the volume of the headspace, it was calculated that the amount of perchloroethylene per single piece was a maximum of 1.0 ng/piece. If we suppose that a patient might be exposed to a maximum of 50 pieces over a lifetime, then the maximum available dose of perchloroethylene from the pieces would be 50 ng. Risk Characterization. A comparison of the available dose (50 ng) to the allowable limit (37 mg/day) indicates an ample safety margin.

23

Ligature Material A manufacturer purchases commercial black fishing line to use as a ligature in a circumcision kit. Because the ligature is not “medical grade,” a cytotoxicity test is routinely conducted as an incoming inspection test. It was assumed that a negative cytotoxicity test would be associated with an acceptable incidence of skin irritation. Hazard Identification. A newly received lot of the fishing line failed the cytotoxicity test. The extraction ration of this material—of indeterminate surface area—was 0.2 g/mL, with a 0.1-mL aliquot of sample extract being applied to a culture dish. Thus, 0.2 g/mL × 0.1 mL = 0.02 g represents a toxic dose of fishing line. Dose-Response Assessment. A titration curve was obtained on the sample extract. If the sample was diluted 1:2, the test was still positive; however, if the sample was diluted 1:4, the test was negative. Thus, 0.02 g/4 = 0.005 g of fishing line, the maximum dose that is not cytotoxic. This value was called the allowable limit of fishing line. Exposure Assessment. Each circumcision kit contained about 12 in. of line, but only about 4 in. of material was ever in contact with the patient. Since an 8-yd line was determined to weigh 5 g, the available dose of fishing line was calculated to be 5 g/288 in. × 4 in. = 0.07 g. Risk Characterization. A comparison of the available dose (0.07 g) with the allowable limit (0.005 g) convinced the manufacturer to reject the lot of fishing line.

Sources of Data Data for calculating the allowable limit of exposure to a material can come from many sources, most of them promulgated by industrial and environmental hygienists and related agencies (Hayes, 1994). Threshold Limit Values (TLVs) are time-weighed average concentrations of airborne substances. They are designed as guides to protect the health and well-being of workers repeatedly exposed to a substance during their entire working lifetime (7–8 hr/day, 40 hr/wk). TLVs are published annually by the American Conference of Governmental Industrial Hygienists (ACGIH, 1986). Biological Exposure Indices (BEIs) are also published annually by ACGIH. These are the maximum acceptable concentrations of a substance at which a worker’s health and well-being will not be compromised. Other published guides include Workplace Environmental Exposure Levels (WEELs), from the American Industrial Hygiene Association (1980); Recommended Exposure Limits (RELs), from the U.S. National Institute for Occupational Safety and Health; and Permissible Exposure Limits (PELs), from the U.S. Occupational Safety and Health Administration. In the United States, PELs have the force of law. Another important limit measurement, Short-Term Exposure Limits (STELs), are defined as the maximum concentration of a substance to which workers can be exposed for a period of up to 15 min continuously, provided that no more than four excursions per day are permitted, and with at least 60 min between exposure periods. The STEL allows for short-term exposures during which workers will not suffer from irritation, chronic or irreversible tissue damage, or narcosis of sufficient degree to increase the likelihood of injury, impair self-rescue, or materially reduce work efficiency. Some substances are given a “ceiling”—an airborne concentration that should not be exceeded even momentarily. Examples of substances having ceilings are

1

Regular

Others

2

Body substances; connections to AMD≥IIa

Regular

3

Regular

Filtration; centrifugation; gas or heat exchange

Channeling or Biological or storing chemical substances modification for of liquids introduction for infusion into the body

Noninvasive devices Body orifices

Transient use

Short-term use

Long-term use, implantable devices

Surgically invasive devices

Invasive devices

4

5

6

7

8

9

10

Immediate danger to heart, respiration, CNS; Ionizing radiation including control monitoring

Regular

11

Potentially hazardous (substances, part of body, mode of applications)

Regular

12

Therapeutic Diagnostic Devices for Others devices for devices to administration administration supply energy, or removal of or exchange vital physiological substances to of energy processes; radioor from the pharmaceutical body imaging Illumination Regular of human body in visible spectrum

Additional rules for active devices

Mechanical Reusable Transient barrier; use; ENT surgical compression; short-term instruments absorption of exudates Regular Short-term Regular Regular Placed in Regular use ENT, teeth long-term Connection to AMD ≥ IIa Wounds with Long-term Ionizing Ionizing Regular Potentially breached radiation; radiation; hazardous use dermis, biological (nature, chemical healing by effect; change density, site secondary absorbed; (except in of energy); intent potential teeth); Class IIb ATD hazard of medicine monitor administration medicine control delivery development system Heart; CCS Heart; CCS; Heart; CCS; CNS; CNS; biological biological effect; effect; absorbed absorbed; chemical change; medicine administration

Contact with injured skin

An Overview of the Classification of Medical Devices (Rules 1–12)

Abbreviations: AMD, Active medical device; ATD, Active therapeutic device; CCS, Central circulatory system; CNS, Central nervous system; ENT, Ear nose and throat.

Rule

III

IIb

IIa

I

Class

Table 2.18

Wounds with breached dermis, healing by secondary intent

Regular

Long-term use

Short-term use ENT, long-term

Body orifices

Biological effects

Regular

Transient use

4

5

7

Heart; CCS; CNS; biological effects

Regular

Short-term use

8

Long-term use, implantable devices

Abbreviations: ATD, Active therapeutic device; CCS, Central circulatory system; CNS, Central nervous system; ENT, Ear nose and throat.

3

6

Filtration; centrifugation; gas or heat exchange Regular

Contact with injured skin

Rule

2

Body substances

Channeling or Biological or storing chemical substances modification for of liquids for introduction infusion into the body

Surgically invasive devices

Invasive devices

Heart; CCS

1

Others

Noninvasive devices

Reduced Scheme for Combined Characteristics of Active and Nonactive Devices

III

IIb

IIa

I

Class

Table 2.19

9

Potentially hazardous (nature, density, site of energy); Class IIb ATD monitor control development

Regular

10

11

12

Regular

Devices for Others administration or removal of substances to or from the body

Immediate danger to Potentially heart, respiration, hazardous CNS; Ionizing (substances, radiation part of body, including control mode of monitoring applications)

Therapeutic Diagnostic devices devices for to supply administration energy, vital or exchange physiological of energy processes; radiopharmaceutical imaging Illumination of human body in visible spectrum Regular Regular

Additional rules for active devices

26

Safety Evaluation of Medical Devices

certain irritants whose short-term effects are so undesirable that they override consideration of long-term hazards.

Uncertainty Factors An uncertainty factor is a correction that is made to the value used to calculate an allowable limit. It is based on the uncertainty that exists in the applicability of the data to actual exposure conditions. Typically, uncertainty factors range in value from 1 to 10. For example, a correction factor of 10 might be applied for data obtained in animals rather than humans, or to allow for a different route of exposure. In other words, for every property of available data that is different from the actual application, a correction factor of between 1 and 10 is applied. If our first example had been of a small amount of data obtained in animals by a different route of exposure, an uncertainty factor of 1000 might be applied.

Safety Margins A safety margin is the difference or ratio between the allowable limit (after correction by the uncertainty factor) and the available dose. How large does a safety margin need to be? Generally, a safety margin of 100× or more is desirable, but this can depend on the security of the risk under consideration, the type of product, the business risk to the company, and the potential benefits of product use.

REFERENCES AAMI (2006). AAMI Standards: Vol. 4, Biological Evaluation of Medical Devices. AAMI, Arlington, VA. AAMI (1998). AAMI Standards: Reduced Devices—Risk Management— Part 1: Applications. AAMI/ISO 14971-1, AAMI, Arlington, VA. ACGIH (1986). Documentation of the Threshold Limit Values for Substances in Workroom Air, 5th edn. American Conference of Governmental Industrial Hygienists, Cincinnati, OH. AIHA (1980). Hygienic Guide Series, Vols. I and II. American Industrial Hygiene Association, Akron, OH. Brown, R.P. and Stratmeyer, M. (1997). Proposed Approach for the Biological Evaluation of Medical Device Materials. In: Proceedings of the Medical Design and Manufacturing East 97 Conference and Exposition. Canon Communications, Santa Monica, 205-9–205-18. CDRH (1992). Regulatory Requirements for Medical Devices: A Workshop Manual. Center for Device and Radiological Health, HHS Publication FDA 92-4165, August 1992. CDRH (1995a). Premarket Notification (510(k)) Guidance Document for Contact Lens Car Products. Center for Device and Radiological Health, Food and Drug Administration, Washington, D.C. CDRH (1995b). Testing Guidelines for Class III Soft (Hydrophilic) Contact Lens Solutions. Center for Device and Radiological Health, Food and Drug Administration, Washington, D.C. CDRH (1995c). Draft Guidance for the Content of Premarket Notifications for Menstrual Tampons. Center for Device and Radiological Health, Food and Drug Administration, Washington, D.C. CFR (1992). FDA’s Policy Statement Concerning Cooperative Manufacturing Arrangements for Licensed Biologics. Federal Register, 57: 55544, November 25, 1992. Council Directive 93/42/EEC of 14 June 1993 Concerning Medical Devices. Official J Eur Communities, 36 (July 12): 1, 1993. Ecobichon, D.J. (1992). The Basis of Toxicology Testing. CRC Press, Boca Raton.

European Committee for Standardization (1991). CEN Annual Report 1991. Brussels. FDLI (1995). Compilation of Food and Drug Laws. Volumes I and II, Volume III (1996), Supplement (1998), Food and Drug Law Institute, Washington, D.C. “Final Draft Guidelines on Medical Device Classification,” MEDDEV 10/93, Brussels, European Commission, October 1993. Food and Drug Administration (1987). Good Laboratory Practice Regulations: Final Rule. Federal Register, Part VI, Vol. 52, No. 172. September 4, 1987. Food and Drug Administration (1995). EPA Bluebook Memorandum #G95: Use of International Standard ISO-10993, Biological Evaluation of Medical Devices Part I: Evaluation and Testing. Food and Drug Administration, Washington, DC. Fries, R.C. (1999). Medical Device Quality Assurance and Regulatory Compliance. Marcel Dekker, Inc., New York. Gad, S.C. and Chengelis, C.P. (1998). Acute Toxicology. Academic Press, La Jolla, CA. Gad, S.C. (2001). Regulatory Toxicology, 2nd Edn. Taylor and Francis, London, UK. Gad, S.C. and Taulbee, S. (1996). Handbook of Data Recording, Maintenance and Management for the Biomedical Sciences. CRC Press, Boca Raton, FL. Goering, P.L. and Galloway, W.D. (1989). Toxicology of medical device material. Fund. Appl. Toxicol., 13: 193–195. The Gray Sheet (1992). EC ‘Medical Devices’ Directive Slated for Adoption in Mid-1993, EC Commission Official Says: CEN Estimate Development of 92 Standards for Directive, M-D-D-1 Reports. The Gray Sheet, October 12, 1992. The Gray Sheet (1996a). European Union Class III Device Approvals Average 240 Days or Less, HIMA Survey Says: Study Release Intended to Bolster Support for FDA Reform Legislation. The Gray Sheet, February 26, 1996, pp. 7–8. The Gray Sheet (1996b). FDA 510(k) Average Review Time for Fiscal 1996 Projected to Be on Par with FY 95 Figure of 137 Days: PMA Average Review Time Expected to Drop to 250 Days. The Gray Sheet, March 25, 1996. Haindl, H. (1997) CE Marking via Self-Declaration. Medical Device and Diagnostic Industry, Sept, 1997, 86–90. Hayes, A.W. (1994). Principles and Methods of Toxicology (3rd edn.). Raven Press, New York, pp. 26–58. Heller, M.A. (1999). Guide to Medical Device Regulation, Volumes 1 and 2. Thompson Publishing Group, Washington, DC. ISO (various dates). Biological Evaluation of Medical Devices. ISO 10993, parts 1–12 (Geneva: International Organization for Standardization, various dates). ISO(1992). Biological Evaluation of Medical and Dental Materials and Devices. ISO, Brussels. ISO (1996). Risk and Hazard Assessment of Medical Devices. ISO 14538, ISO, Brussels. Japanese Ministry of Health and Welfare (1995). Guidelines for basic biological tests of medical materials and devices, Notification No. 99. Japanese Pharmacopoeia XIII (1996). Kahan, J.S. (1995). Medical Devices—Obtaining FDA Market Clearance. Parexel, Watham, MA. Stark, N.J. (1997). Case Studies: Using the World Literature to Reduce Biocompatability testing. Proceedings of the Medical Design and Manufacturing East 97 Conference and Exposition. Canon Communications, Santa Monica, 205-1–205-7. Stark, N.J. (1998) Conducting Health-Based Risk Assessment of Medical Materials. Medical Plastics and Biomaterials, Sept./Oct., 18–25. USP (2007). Biological tests—plastics. The United States Pharmacopoeia, XXX Edition. United States Pharmacopoeial Convention, Inc., Rockville, MD., pp. 1235–1238.

3 Road Map to Test Selections

Determining what testing is required for the development and approval of a new medical device can be a complex issue. This is even more the case after the issue of when to perform necessary tests is factored in. Essentially, if one or more clinical trials (with enabling IDE) is part of the development scheme, exactly how much biocompatability work must be conducted prior to the IDE filing is not a fixed or well-defined list. Post approval, one must determine what ongoing testing is required to ensure continued safety of the product. Understanding the complexities requires careful consideration of some key concepts.

not planned, such as using elastic bandages to cover wounds. Such is especially (but not exclusively) the care for children and the elderly, who are more likely to be susceptible to adverse effects. It should also be kept in mind that, though many electronic devices (disposable gloves and syringes, for example) are intended to be singleuse disposables, in poorer cultures this frequently may not be the case. One cannot guard against every—or even most unusual—device usage, but should exercise some consideration of what the most likely misuses are.

MATERIALS, COMPONENTS, AND PRODUCTS KEY CONCEPTS

What is actually sold for use by or on the end use consumer (the patient) is what is regulated by the various government agencies. This is the product, which must be evaluated for biocompatibility in conformance with applicable guidelines, in the form or forms (sterilized or unsterilized) that it is intended to be sold. But products are frequently composed of components. Simple examples are a disposable syringe (needle, barrel, plunger, lubricant, and stopper) or a surgical prep set (scrub, disinfectant, razor, etc.). Changing a component can significantly alter the biocompatibility of a product. And certain components, by the nature of both their composition and exposure to patients, are more likely to present biocompatability problems. An example is the common disposable plastic syringe, of which billions are used each year. For the syringe, the most likely problem component is the stopper— the flexible piece at the end of the plunger. The stopper is most commonly made of natural rubber, and has direct contact with fluids entering the body (and frequently a fluid path). This is the most common problem component for a syringe. Components, of course, are manufactured from or composed of materials. Materials (polymers, elastomers, steel, etc.) are the fundamental starting point for development of a device, and are very frequently not produced by the device (or component) manufacturer, but are rather provided by an outside vendor. Almost all biocompatability problems (the exceptions being due to sterility, pyrogenicity, sterilization, and cleanliness) for devices are due to the materials used in a device. Table 3.2 provides a concise list of material-based considerations for safety of a device.

There are ten major categories on consideration in evaluating and establishing the safety of a medical device, and in so doing defining what testing must be performed. These are presented as a list in Table 3.1, and will each be considered in detail in this chapter.

CONDITION OF USE The starting point for evaluating the safety of a near (or potential) medical device (or material for use in devices) must be understanding both how it is intended to be used (which governs the type of contact it will have with the end use consumers, i.e., patients, and therefore the areas of potential risk and the applicable regulations) and how it is likely to be used (or misused). Intended use starts with developing an objective statement of what purpose the device is to serve, and therefore how it is to be in contact with the patient (skin/body surface, only body cavity, indirectly with a fluid path (such as and most commonly the blood stream) within the body) and for how long there is to be patient contact. The categories for type of contact are drawn from the nation’s regulatory guidelines presented earlier, but actual devices may fit in several categories. It is also important to know more details of the contact (such as what body cavity contact is with— mouth, nasal, vagina, anal, etc.). The duration of cumulative patient exposure in use should also be established. This should be the cumulative duration for any patient, and not just the single time/use duration. (That is, if a device is to be used 5 min a day each day for a week, it should be considered to have an approximately 35 min cumulative patient exposure). For most devices, the intended use for any one patient is a single time. But if the device (say a glove) is used by a health care provider, over the course of a day, cumulative exposure can be extensive. One must also consider unintended uses or expected abuses of the device. People use devices in ways that are

CHEMICAL AND PHYSICAL PROPERTY CONSIDERATIONS Engineers involved in the design and development of new medical devices are primarily concerned with the physical properties essential for the proper functioning of the device. Accordingly, the most important aspects of materials being 27

28

Safety Evaluation of Medical Devices

Table 3.1

Key Conceptsa in Medical Device Safety Assessment

1. Condition of Use 2. Materials/Components/Products 3. Chemical and Physical Property Considerations 4. Factors of Influence 5. Prior Knowledge 6. Types and Uses of Tests 7. Reasonable Man 8. Qualification vs. Process Control 9. Tiers of Concern: Consumer/Health Care Provider/Manufacturer 10. Sterilization a

These also could be thought of as areas of consideration.

used in device construction are its physical properties. Toxicologists and others responsible for device safety (the subject of this volume) are primarily concerned with the chemical nature and properties of materials used in devices, but must also be conversant with physical property considerations. For that reason, this section presents a primer on the chemical nature and chemical and physical properties of materials used in devices. Physical or mechanical properties can also be important for biocompatibility. Mechanical insult is a localized, nonspecific, detrimental mechanical interaction between a device and the contacting tissue. It may be caused by rubbing, crushing, occluding, stripping, or penetrating the tissue. Rubbing may cause layers of tissue to separate from one another, resulting in the formation of blisters or bare underlying dermal layers. Crushing, or pressure damage, may result in form tissue swelling beneath an inelastic device. Occlusive materials may encourage moisture and bacteria to accumulate beneath them. Tissue may be stripped when adhesive materials are removed, tearing the epidermis away from the dermis. Tissue penetration, due to incision or puncture, will cause cells to be ruptured and physically separated from each

Table 3.2

Raw Material Characterization

Chemical Characterization • list materials • list potential extractables or leachables from materials • physiochemical tests, USP water and isopropanol extracts • international pharmacopoeial tests • infrared analysis document polymer identification • chromatographic characterization molecular weight distribution additive and/or extract analysis • trace metals • specific gravity • moisture content Physical Characterization • hardness • surface characterization • color, opacity or haze • strength properties tensile/elongation flexure compression • thermal analysis • viscosity, melting point, refractive index Biological Characterization • [nothing listed]

other. Indigestible particulate matter may be shed into or left embedded in the tissue. Microscopically, mechanical insult, like chemical insult, is characterized by inflammation. Such insults reflect tissue to material interface interaction, for which the nature of the device surface plays a significant role. Macroscopic responses to mechanical injury are usually obvious and include dermatosis, callus, granuloma, or cyst formation. Histologically, responses to mechanical injury may include tissue infiltration by phagocytic macrophages and foreignbody giant cells, and by fribroblast, whose function is to deposit collagen and wall off the offending material within a cyst. Whereas the tissue effects of chemical insults are generally symmetrically distributed around the sample as the offending chemical is extracted, those of mechanical insult are generally asymmetrically distributed according to the geometry of the material. Mechanical failure includes device failure that causes injury to the organism or loss of a life support mechanism. Mechanical failure is not generally considered to be an issue of biocompatibility, although certainly it is an issue of device safety. The majority of components of medical devices are constituted from a small number of categories of materials. The important categories of materials in device formulation and construction are the following: water, stainless steels, polymers, elastomers, silicones and natural fibers (cotton and wood pulp, primarily). Each of these needs to be considered in turn. Additionally, biological derived materials are seeing increased utilization in devices (Kambric et al., 1986), but are of such diverse nature that is not currently possible to adequately overview them in this book. These are interactions which occur between material used in a device and the organism (“host”) that it has contact with. Device materials having systemic contact with a host (“biomaterials”) may be degraded by the host by a number of chemical means which should be kept in mind when considering the use of any of the materials described in this section. These pathways of chemical degradation of biomaterials include: Hydrolysis (acid, base, neutral aqueous media) Oxidation (corrosion, chain cleavage) Thermolysis Photooxidation Specific enzyme—catalyzed hydrolysis or oxidation Attack by complex media (culture media, serum, blood, gastric juices, urinary fluids, phagocyte-containing fluid, etc.) Chain cleavage due to mechanical fracture. The chemical basis of toxicity or biocompatibility problems are related to these same processes. In theory the biocompatibility of a material could be assessed by a careful analysis of its chemical composition, but in practice the exact chemical composition is usually either proprietary or unknown and too costly to identify. It is common knowledge that virtually no plastics are pure polymers—all are modified with inadvertent contaminants and intentional additives. Fabrics purchased from the textile industry contain a variety of finish coatings. Natural materials are usually impure and vary from lot to lot. Synthetic materials contain organic residues. Metal alloys contain leachable essential trace elements. Because of these variables, toxicologists usually adopt a different approach to establishing biocompatibility.

Chapter 3:

The interaction of a material with living tissue is assessed in vivo: the material is considered biocompatible with regard to a particular mechanism of injury if there is little or no tissue response; it is considered nonbiocompatible if there is a response. Three factors are important: the rate at which an additive or contaminant leaches out of the material, the effect that the loss of additives has on the material, and the toxicity of the additive or contaminant. When a positive response occurs, the toxicologist must play chemical sleuth to identify its cause and to recommend process or formulation changes to eliminate it. Some of the sources of possible toxic substances in medical device materials are the following. Each source is then discussed separately. • Residual monomers. • Residual solvents. • Degradation products (for which ISO-10993 now provides specific guidance). • By-products from irradiation. • Sterilization residuals. • Formulation additives. • Inadvertent contaminants. • Bacterial endotoxins.

Road Map to Test Selections

29

Polyglycolic acid, used in suture production, is virtually destroyed by irradiation. Most medical polymers decrease in molecular weight as a result of chain scission. Polypropylene and other polymers may undergo chain scission, cross-linking and oxidation. Any material that is irradiation sterilized should be tested for biocompatibility afterward, although most medical polymers remain useful.

Sterilization Residuals Chemical sterilization with ethylene oxide has a long history of use; the main advantages are that the procedure is carried out at low temperatures and the sterilization facility need not deal with radioactive sources. Ethylene oxide, which is itself toxic, also degrades into toxic ethylene chlorohydrin and ethylene glycol. Even after extensive degassing, some materials do not release these toxic molecules. All materials that are to be ethylene oxide sterilized must be tested for toxic residuals. These are discussed in Chapter 13.

Formulation Additives

Materials may undergo degradation during manufacture, sterilization or storage, or after application to or implantation in the body. During manufacture, heat may thermally degrade a material; polyvinyl-chloride is especially susceptible to heat and may release hydrochloric acid, resulting in an autocatalytic unzipping process. During sterilization, polytetrafluorethylene is susceptible to irradiation breakdown, resulting in the release of hydrofluoric acid. Stored materials exposed to light and oxygen may suffer ultraviolet degradation or oxidation. Implanted materials, particularly metals, may corrode or be biologically degraded (ISO, 1997). Stabilizers added to the polymer can protect against degradation. Materials should be tested for degradation and biocompatibility only after both manufacture and a suitable aging period.

Formulation additives include plasiticizers, stabilizers, antioxidants, fillers, catalysts, mold release agents, colorants, antistatic agents, preservatives, and flame retardants. Both synthetic and natural fabrics are apt to have a variety of finish coatings. Plasticizers allow the polymeric chains of a plastic to slide past each other, providing material flexibility. Stabilizers protect the plastic against heat, oxygen, and light. Antioxidants protect it from oxidation. Fillers expand the volume of the plastic and sometimes impart mechanical properties. Catalysts are small molecules that promote polymerization. Mold release agents facilitate the release of a molded part from the mold. Colorants add color and beauty. Antistatic agents prevent electrostatic buildup. Preservatives protect against microbial degradation. Flame retardants slow the release of toxic fumes during extreme conditions of heat or flame. Coatings impart a variety of properties to fabric, such as resistance to mildew or absorbency. Any of these additives may contribute to nonbiocompatibility. To complicate the picture, because the medical device industry uses only a small fraction of the total volume of industrial plastics, few device manufacturers are in a position to specify the exact formulation of the material they will use. (Even so-called medical-grade plastics contain additives, medical grade usually taken to mean that the material has passed certain testing regimens, usually USP grade VI, and that it is processed under clean conditions.) Thus, manufacturers must frequently evaluate industrialgrade materials to obtain the mechanical and chemical properties required in the final product. Fabric devices, like any other dermal contact product, must also be evaluated for biocompatibility. Those that will be laundered between uses present the user with less finishcoating exposure than do disposable devices that bring new fabric with each use. Some finish coats are carcinogenic; others cause allergic reactions or skin irritation. Some are water-soluble and are removed during washing; others are organic based and will remain.

By-Products from Irradiation

Inadvertent Contaminants

Gamma irradiation is becoming an increasingly common method of sterilization; 2–3 Mrad is the usual sterilization dose. It is also used to facilitate cross-linking in certain formulations. As a result, many materials undergo degradation.

Materials may also contain inadvertent contaminants or particulates that are introduced during manufacture. Metal shavings from equipment wear or plastic shards from stamping processes have been known to adulterate devices.

Residual Monomers Polymerization results in a distribution of molecular weights. Although monomers are usually toxic, the toxicity of the polymeric unit generally decreases as the molecular weight increases. Residual monomers result from incomplete polymerization; their concentration can be controlled by carefully regulating polymerization conditions.

Residual Solvents Solvents are often an integral part of manufacturing and may remain behind in fluid materials such as adhesives, adhesive removers, barrier pastes, gels, or lubricants. Some acrylate adhesive systems are now water based, which eliminates the concern for residual solvents. When the solvent is an integral component of the final product, as with adhesive removers and certain barrier pastes, its presence and potential transdermal absorption must be addressed in labeling.

Degradation Products

30

Safety Evaluation of Medical Devices

Table 3.3

Properties of Major Metals Used in Medical Devices

Material: Metals and metallic alloys

Strength

Tensile modulus

Creep modulus

Fatigue life

Lubricity

Water permeability

Water absorption

Biostability

Cobalt chrome alloys Nickel chrome alloys Nitinol alloys (shape memory alloys) Stainless steels Tantalum Titanium and titanium alloys

High High High High High High

High High High High High High

High High High High High High

High High High High High High

Mod Mod Mod Mod Mod Mod

Low Low Low Low Low Low

Low Low Low Low Low Low

High High High Med High high

Bacterial Endotoxins Finally, materials may contain bacterial endotoxins, most commonly introduced through water during manufacture. Contaminating endotoxins are of concern for devices that will be used on anything other than intact skin or mucous membrane. They are readily eliminated by using endotoxinfree water. Issues around the specific primary materials used in device manufacture should now be considered. Note that the characterization of materials is specifically addressed in ISO 10993-14.

SPECIFIC MATERIAL CONSIDERATIONS Water Water, in one way or another, is involved in the production of virtually every medical device. For many devices (particularly diagnostics), it is also incorporated into the device. Yet water tends to be invisible to many considered device biocompatibility and safety. Water’s greatest uses are in cleaning and rinsing devices and their components. Purified water is obtained by distillation, ion-exchange treatment, reverse osmosis, or other suitable processes. Such water is prepared from source material complying with the regulations of the US Environmental Protection Agency (USEPA) for drinking water. Water can be a problem in device safety and biocompatibility generally due to the presence of things in it which do not comply with USEPA and other regulations (such as microorganisms, endotoxins, pesticides, organics, other pyrogens, and heavy metals), as well as contaminants from other devices or components which may have been previously rinsed in the water (such as latex from gloves or stoppers). Water is usually a source of problems in production of devices, and not in the development stage. Specific issues and standards associated with heavy metals will be addressed in Chapter 13.

chromium. Table 3.4 presents the ASTM specifications of the two grades of stainless steel which are commonly used for medical devices. Stainless steel is rarely a biocompatibility problem in medical devices. In some long-term implants, there can be concerns about reaction of the alloy with body fluids, but even this is unusual. The other major metals are primarily used in surgical implants, particularly orthopedic (bone) prosthesis. These are primarily titanium and cobalt alloys, though some more exotic metals are also seeing use. Gold, palladium, and platinum, for example, are plated onto the surface of stents to provide radiopacity for placement and monitoring. Silver is used as an antimicrobial. Unlike stainless steel, not all of these metals are biological inert. The compositions and properties of the other major metal alloys employed are summarized in Tables 3.5–3.7.

Polymers Polymers are macromolecules formed by the chemical bonding of five or more identical units called monomers. In most cases the number of monomers is quite large (3500 for pure cellulose), and often is not precisely known. In synthetic polymers, this number can be controlled to a predetermined extent, e.g., by shortstopping agents. (Combinations of two, three, or four monomers are called, respectively, dimers, trimers, and tetramers, and are known collectively as oligomers. Such oligomers are not polymers.) A partial list of polymers by type includes the following (Billmeyer, 1971): 1. Inorganic: siloxane, sulfur, chains, black phosphorus, boron-nitrogen, silicones 2. Organic a. Natural i. Polysaccharides: starch, cellulose, pectin, seaweed gums (aga, etc.), vegetable gums (arabic, etc.) Table 3.4

Chemical Composition of Stainless Steel Composition (%)

Metals A variety of metals sees significant use in medical devices, though with the exception of stainless steel their use in patient contacting situations is largely limited to implants. The uses are a reflection of the properties of the various metals, as summarized in Table 3.3.

Stainless Steel Stainless steel is by far the most common metal used in devices. This ferrous alloy is widely used in surgical instruments, wire sutures, needles, screws, and implant parts where great strength is needed. Steels are, by definition, alloys of iron containing 0.002% to 1.5% carbon, while stainless steel contains high percentages (~10–25%) of

Element Carbon Manganese Phosphorus Sulfur Silicon Chromium Nickel Molybdenum Nitrogen Copper Iron Source: ASTM F56, 1986.

Grade 1

Grade 2

0.08 max 2.00 max 0.030 max 0.030 max 0.75 max 17.00–19.00 12.00–14.00 2.00–3.00 0.10 max 0.50 max Balance

0.030 max 2.00 max 0.030 max 0.030 max 0.74 max 17.00–19.00 12.00–14.00 2.00–3.00 0.10 max 0.50 max Balance

Chapter 3:

Table 3.5

31

Road Map to Test Selections

Cobalt-Based Alloys Wrought CoNiCrMo MoWFe

Condition Source Density (g/cm3) E (tensile) (GPa) Hardness (Hv) s0.2% (MPa) sUTS (MPa) Elong. (min%)

Cast wrought CoCrMo CoNiCr–

Wrought CvoCrMo

AN [1,3] 9.15 — 240 390 880 30

AN [4] — — — 240-450 795-1000 50

AN [1,2] 7.8 230 300 455 665 10

Wrought CoNiCr– MoWFe

Wrought CoNiCrMo AN [5] — 230 — 275 600 50

Wrought CoCrMo

AN [1,3] 9.15 — 450 1000 1500 9

CW [4] — — — 1585 1795 8

Abbreviations: AN: annealed; CW: cold worked; —: unavailable. Sources: [1]: BSI 3531 (Part 2, Sec. 4-5, Amend. 2, 1983); [2]: ASTM F90-82; [4]: ASTM F562-84; [5]: ASTM F563-83.

Table 3.6

Titanium and Titanium-based Alloys

Condition Source Density (g/cm3) E (tensile) (GPa) Hardness (Hv) s0.2% (MPa) sUTS (MPa) Elong. (min%)

Ti type 4

Ti6A14V

AN [1,2] 4.5 127 240–280 430–365 550–575 15

AN [1,3] 4.4 127 310-350 830 900 8

Ti5A12.5Fe

Ti6A14V

Ti6A17b

HF [6] 4.4 127 — — — —

HF [4,5] 4.52 105 400 800–900 900–1000 10–12

HF [7] 4.45 — — 815 965 16

To5A12/5Fe [7] 4.45 — — 900 985 13

Abbreviations: AN: annealed; HF: hot forged; —: unavailable.

Table 3.7

Other Metals and Alloys

Material Condition Source Density (g/cm3) E (tensile) (GPa) Hardness (Hv) s0.2% (MPa) sUTS (MPa) Elong. (min%)

Ta

Ta

Pt

Pt10Rh

Pt10Rh

W

16.6 186 — 140 205 20–30

AN [1,2] 16.6 186 — 345 515 2

CW [1,2] 21.5 147 38–40 — 135–165 35–40

AN [1] 20 — 90* — 310 35

AN [1] 20 — 165* — 620 2

75%CWSN [1] 19.3 345 225 — 125–140 ~0

Abbreviations: AN: annealed; CW: cold worked; SN: sintered bar; *: Brinell hardness; —: unavailable.

ii. Polypeptides (proteins): casein, albumin, globulin, keratin, insulin, DNA iii. Hydrocarbons: rubber and gutta percha (polyisoprene), also called elastomers b. Synthetic i. Thermoplastic polymers: nylon, polyvinyl chloride, polyethylene (linear), polystyrene, polypropylene, fluorocarbon resins, polyurethane, acrylate resins ii. Thermosetting polymers: polyethylene (crosslinked) phenolics, alkyds, polyesters c. Semisynthetic cellulosics (rayon, methylcellulose, cellulose acetate) and modified starches (starch acetate, etc.). For most devices, we are concerned only with the synthetic organic polymers. The principal class of natural polymers of concern, the elastomers, will be considered later in this chapter along with the synthetic members of this

class. The chief class of inorganic polymers of concern, the silicones, will likewise be considered later in this chapter. Polymers can be categorized in a number of ways, (Haslam et al., 1972; McMurrer, 1985). Homopolymers, for example, consist of only one repeating monomer unit. Table 3.8 lists the most commonly encountered homopolymers. Figure 3.1 provides the structures of some typical monomers, while Figures 3.2–3.6 provide structural illustrations of some homopolymers. Copolymers are produced by the simultaneous polymerization of two or more dissimilar molecules. Examples include polyvinyl acetate (Fig. 3.7), polyesters (Fig. 3.8), and polymides (Fig. 3.9). Synthetic elastomers (such as SBR synthetic rubber, made from styrene and butadiene) are also copolymers. This pattern continues with the terpolymers (such as ABS, shown in Fig. 3.10), which consist of three different monomers.

32

Safety Evaluation of Medical Devices

Table 3.8

Commonly Used Homopolymers in Medical Devices

Polyethylene Polypropylene Polybutylene Polyvinylchloride

Polystyrene Polychloroprene Polyacrylates Polytetrafluoroethylene

Polysiloxanes Polysulfones Polyamides

CH2

CH

CH2

CH

CH3

Source: Autian, 1980.

CH3

CH2

CH

CH3

Figure 3.2 Structural diagram of polypropylene, a typical homopolymer, along with that of its monomeric unit (propylene).

The principal concerns with the biocompatibility of polymers are additives, residual monomers, and contaminants which are leachable in the body. As Table 3.9 shows, polymers themselves generally have very low toxicities. Partially as a reflection of their high molecular weights, true polymers themselves are not generally absorbed into the body, are not irritating, and are not sensitizers. Residual monomers, additives, and contaminants are a very different story. Each of these can be quite significant biologically. Residual monomers are those remaining individual building block units in homopolymers, copolymers, terpolymers, etc. (as diagrammed structurally in Figs 3.1–3.10) that are not successfully incorporated into the plastic during the synthesis process. Technically, we should also include dimers, trimers, and other small-chain fragments which are left in the polymer mass but are not chemically bound to it. Many factors help determine how much residual monomer will be left in a polymer, and how available such residuals are to a surrounding biological matrix. And some of the monomers are quite active biologically. When we test a plastic for biocompatibility, biologically available (leachable) residual monomers are a significant part of our concern. Examples of toxic monomers (and their principal toxicities) which can be found in polymers include the following: Acrylonitrile: human carcinogen (liver, brain) Vinyl chloride: human carcinogen (liver) Formaldehyde: animal carcinogen (nasal) Methylene dianline: suspect human carcinogen. A wide variety of other chemical entities are specifically incorporated into plastics to achieve desired goals of structure, performance and processing ease. Table 3.10 presents a short list of the major categories of additives. Such additives can be quite significant biologically, and may be very biologically available. A historical example is diethylhexylphthalate (DEHP), a once widely used plasticizer that was found both to be an animal carcinogen and to

acrolonitrile

H2C

CH—CN

vinyl chloride

H2C

CH—CI

butadiene

H2C

CH—CH

methyl methacrylate

H2C

CH—COOCH3

migrate readily from plastic bags and tubing to the blood and intravenous solutions they contained (Peck et al., 1979; Sasakawa and Mitomi, 1978). Figure 3.11 presents the results of one study of DEHP migration into blood products stored in bags containing DEHP. During the processes of polymer synthesis and forming of the components and products, materials (such as solvents and mold release agents) can contaminate the polymer product (Petersen et al., 1981). The result of the additives and contaminants being in plastic is that a range of toxic materials may be leached from many plastics. Table 3.11 presents a short list of some of the more significant of these. For the interested reader, Kroschowitz (1990) and Dumitriu (1994) are excellent texts on polymers, particularly as they are utilized in medical devices.

Elastomers The classical definition of an elastomer is a member of a class of synthetic thermosetting polymeric substances possessing rubber-like qualities (such as the ability to regain shape after deformation, to be stretched to at least twice its original length, and having toughness and resistance to weathering and chemical attacks). For our purposes, however, the natural rubbers will also be included under the term. The chemical requirements for useful elastomers include high molecular weight, flexibility of the polymer chain, and a controlled degree of polarity and structural irregularity. These features impart cohesiveness and a lack of molecular order in the undeformed state, although natural and synthetic rubbers generally take on some aspects of crystallinity at very low temperatures or when they are highly extended. Natural rubber is a natural organic substance obtained from the sap of certain trees. Rubber is produced from

CH2 CH2 CH CH3

n

CH3

Figure 3.1

Molecular structures of four typical isomers.

Figure 3.3 Structure of polypropylene, a typical homopolymer, rendered as a molecular model.

Chapter 3:

CH2

CH

Made from: styrene

CH2

CH

CH2

CH

Road Map to Test Selections

33

CH

CH2 CH

CH2

Figure 3.4 Structural diagram of polystyrene, a typical homopolymer, along with that of its monomeric unit (styrene).

a wide variety of plants, especially Hevea brasiliensis (family Euphorbiaceae), a tall softwood tree native to Brazil, but also from species of Mabea, Manihot, Sapium, Ficus, and others. The milky liquid (latex) found in the inner bark and obtained by tapping (cutting) the bark is coagulated; the oldest and simplest technique is by smoking over a fire. Plantation rubber, widely grown in Southeast Asia (particularly Malaysia), comes from trees grown from selected seeds, with trees spaced well apart; tapping similar to the method used for wild trees, but coagulation is accomplished chemically. Most rubber used in medical devices is still of the natural variety due to economics and the better resistance of the natural product to heat. Recent concerns about latex allergies, however, have increased the move towards synthetic products. Synthetic rubber (or the proper elastomers) comprises a family of elastic products derived chiefly from petroleum and alcohol, some of which closely resemble natural rubber and some of which have quite different physical properties. Research on the chemistry of natural rubber led in the 19th century to the isolation of isoprene, which was reconverted into a rubber-like substance by distillation. This and other processes by which long chains of molecules were created (called polymerization) attracted continued research into the early 20th century.

CH2 CH3 n

Cl

n

Figure 3.6 Structure of polyvinyl chloride (PVC), a typical homopolymer, rendered as a molecular model along with the structure of the monomeric unit (vinyl chloride).

Among the most important synthetics are the butadiene– styrene copolymers; the various butadiene copolymers; the chloroprene polymers (neoprene); the polysulfide polymers (Thiokol); cis-1,4-polyisoprene, ethylene–propylene terpolymers (EPDM rubber); polyurethane rubber; the isobutylene polymers (butyl rubbers); the silicone rubbers produced by blending silicones with natural and synthetic rubbers; and plasticized vinyl chloride and vinyl acetate polymers and copolymers. These can be cross-linked with sulfur, peroxides, or similar agents. The term was later extended to include uncross-linked polyolefins that are thermoplastic; these are generally known as TPO rubbers. Their extension and retraction properties are notably different from those of thermosetting elastomers. Many important elastomers are copolymers in which the main molecular chain is composed of carbon atoms; they are usually based on butadiene and other organic compounds obtainable from petroleum: in Buna S, the second component is styrene; in Buna N, acrylonitrile. Butyl rubber, used in pneumatic inner tubes, is a copolymer of isobutylene and isoprene. Elastomeric materials in which the main polymer chain contains atoms other than carbon include the polysulfide rubbers (Thiokols) and the silicones. We will consider the silicones as a separate class of compounds. Synthetic rubbers, like natural rubbers, can be toughened by vulcanization and improved and modified for special purposes by reinforcement with other materials. Elastomers are used as closures in vessels; stoppers in syringes; and in gloves, tubing, condoms, and in a host of ways in other devices. The degree of curing of elastomers is particularly important in determining their biocompatibility. Natural rubbers are natural products, of course, and therefore subject to significant lot-to-lot variability in composition.

Silicones

Figure 3.5 Structure of polyethylene, a typical homopolymer, with a structural diagram of its monomer, ethylene.

Silicones (organosiloxane polymers) are a large group of siloxane polymers based on a structure of alternating silicon and oxygen atoms with any of a variety of organic radicals attached to the silicone, such as shown in Figure 3.12 (Hardman and Torkelson, 1986; Rochow, 1951). Silicones are very widely used in devices and the pharmaceutical industry, and in such uses the majority of the substituent organic groups are methyl (McGregor, 1953).

34

Safety Evaluation of Medical Devices

CH2

CH

CH2

CH

O

CH2

CH

CI

C

CH2

CH

O

CI

C

O

CH3

O

CH3 O

Made from: H2N

vinyl acetate

CH

H2C

vinyl chloride

OC CH

CH3

CI

Figure 3.7 Structural diagram of vinyl acetate, atypical copolymer, along with the structures of its two constituent monomers (vinyl acetate and vinyl chloride).

O

(CH2)X

O

C

(CH2)y

C

O

O

(CH2)X

O

O

C

(CH2)y

O

Made from:

Figure 3.8

glycols

HO

di-acids

HOOC

(CH2)X

OH

(CH2)y

COOH

Structural diagram of a polyester, a typical copolymer, along with the structures of its two constituent monomers (glycols and di-acids).

NH

(CH2) X

NH

C

(CH2) y

C

NH

(CH2) X

C

O

O

O

NH

Made from:

Figure 3.9

diamines

H2N

di-acids

HOOC

(CH2) X (CH2) y

NH2 COOH

Structural diagram of a polymide, a typical copolymer, along with the structures of its two constituent monomers (diamines and di-acids).

Silicones have an unusual array of properties. Chief among these are thermal and oxidative stability and physical properties little affected by temperature. Other important characteristics include a high degree of chemical inertness, resistance to weathering, good dielectric strength, and low surface tension. As the general formula implies, the molecular structure can vary considerably to include linear, branched, and cross-linked structures. These structural forms and the substituent (R) groups provide many

combinations of useful properties that lead to a wide range of commercially important applications. Silicones include fluids, resins, and elastomers. Many derived products, e.g., emulsions, greases, adhesives, sealants, and coatings, have been developed for a large variety of uses in the device industry. Silicones can be liquids, semi-solids (gels), or solids depending on the molecular weight and degree of polymerization. Viscosity ranges from less than 1 to more than

Chapter 3:

CH2

CH

CH2

CH

CH

CH2

CH2

Road Map to Test Selections

35

CH

CN

y

x

z

Made from:

Figure 3.10 and styrene).

acrylonitrile

H2C

CH

CN

butadiene

H2C

CH

CH

styrene

H2C

CH

CH2

Structural diagram of ABS, a typical terpolymer, along with the structures of its three constituent monomers (acrylonitrile, butadiene,

1 million centistokes. Polymers may be straight-chain, or cross-linked with benzoyl peroxide or other free radical initiation, with or without a catalyst. They have very low surface tensions, extreme water repellency, high lubricating properties, excellent dielectric properties and are permeable to gases and highly resistant to attack and decomposition by biological systems (Heggers et al., 1983). Silicones are used as adhesives, lubricants, protective coatings, wetting agents, mold release agents, surfactants, foam stabilizers, surgical membranes and implants, gaskets, and tubing in the device industry. They have been used as implantable devices since the late 1950s, and in recent years (in the form of prosthetic breast implants) have become of concern for purported health effects and have acted as potent litogens (see the discussion of this aspect in Chapter 17).

Fibers Fibers (both natural and synthetic) are also widely used in medical devices as sutures, connectors, and binding and absorbent materials. Natural fibers that are commonly

employed as silk (typically as suture material), linen, and cotton (primarily for its absorbent properties in bandages, tampons, and such). The natural fibers do not usually cause biocompatibility problems, if properly prepared, cleaned, and sterilized, though environmental contamination of the fiber is a possibility. Synthetic fibers can be of either a durable or shortlived nature. Nylon and rayon are used for durable applications, with rayon being used as an absorbent material in the place of cotton. For suture use, absorbable materials have been developed for use as sutures, with the advantage that an additional surgical procedure is not required to remove them after wound healing is achieved (Schmitt and Polistina, 1967; Elbert et al., 1971; Frazza and Schmitt, 1971; Katz and Turner, 1970). Figure 3.13 presents structures of commonly used nonabsorbable suture materials, and Figure 3.14 those of common absorbable suture materials. There has recently been some (seemingly unfounded) concern about the use of rayon and some other synthetic fibers due to possible contamination with diaxanes.

Ceramics Table 3.9

Ceramics are products manufactured by the action of heat on earthy raw materials (i.e., those that are neither metallic

Oral Lethalities of Common Polymers

Polymer Polyethylene Polypropylene Polychloroprene latex Chlorosulfonated polyethylene Polyvinyl acetate Polyacrylonitrile Polyacrylamide Aromatic polyamides Source: Autian, 1980.

Rat LD50 (g/kg BW) >8 >8 >40 >20 >25 >3 >8.2 >7.5

Table 3.10 Plasticizers Lubricants Antioxidants Colorants Emulsifiers Stabilizers Curing agents

Additives Used in Plastics UV absorbers Blowing agents Fillers Release agents Flame and fire retardants Accelerators Antistatic agents

36

Safety Evaluation of Medical Devices

45 40 35

DEHP mg/unit

30 25 20 15 10 5 0 0

7

14

21

28

35

42

STORAGE PERIOD (DAYS)

Figure 3.11 Average amount of DEHP in stored blood products. 䉭, whole blood, 0–5°C; 䉱, plasma fraction, 0.5°C; 䊐, packed red cells, 0–5°C; 䊏, platelet concentrate, 22°C. Source: Miripol and Stern (1977).

or organic), with silicon and its oxides, aluminum oxides, and complex materials known as silicates occupying a predominant position. In physical structure, ceramics may be polycrystalline, glass, combinations of multicrystals with glassy phases, or single crystals. The universal properties of ceramics that account for their enduring utility include mechanical strength in spite of brittleness’ chemical durability (at both normal and elevated temperatures) against deteriorating effects of oxygen, water (liquid or vapor), acids, bases, salts and organic solvents; hardness contributing to resistance against erosion; and the ability to be combined with metals and other materials to make uniquely effective devices and device parts. The forming of artificial ceramics involves particle technology, including size reduction (commination), measurement, and separation; blending and packing of particles; surface chemistry and physics;

Table 3.11

Identified Toxic Materials in Polymers

Aluminum Acrylonitrile (monomer) Arsenic Benzene Benzoic peroxide Bisphenol A Cadmium Carbon tetrachloride Dibutyl tin Epoxy curing agents Ethylene dichloride Ethylene oxide Formaldehyde Sources: Autian, 1980; Northup, 1989.

Ketones and hydrocarbons Lead Mercaptobenzothiazole Methyl chloride Methylene chloride Methylene dianline Nickel PAHs on carbon black Pyrene Tin Tricresyl phosphate Triphenyl phosphate

rheology, or the flow of more or less plastic materials under pressure; and bonding of particles. Other sciences contribute to the understanding and control of heat treatment. Ceramics came into being because of the useful rheological properties of the clay–water system, with its capability of being shaped, whether as a paste or as a fluid suspension or as a damp, crumbly powder. Modern industrial ceramics often replace clay with organic or inorganic additives to make the raw material easier to form and to provide strength before firing (final heating). Clay-based ceramics undergo both chemical-mineralogical and physical changes during heat treatment to form new phases, including a glassy phase that often constitutes the bond between the grains in the new microstructure. Such heat treatment involves sintering in the presence of a liquid phase. The complex series of occurrences accompanying heat treatments include gas evolution, oxidation, valence changes, chemical reactions, phase transformations, melting, shrinkage, and still other chemical and physical changes. The trend with modern ceramics technology is to simplify such changes by establishing the chemical and mineral changes before forming. Maximum density is sought in the forming batch, and compaction is achieved by applying the principles of particle packing. Then heat treatment brings about sintering in the solid state with a minimum of shrinkage, gas evolution, and attendant flaws, such as warping, cracking, or the formation of unwanted pores. Many significant ceramic products, such as thermal insulation and filters for liquids and gases, require controlled porosity, both in terms of percent by volume and in pore diameters and distributions. In fact, texture compromises the prime distinction between refractories, whitewares, and electronic and glass-ceramics that may otherwise be identical in chemical and mineral composition.

Chapter 3:

CH3

OSi

CH3

O

CH3

Figure 3.12

CH3

Si

OSi

CH3

n CH3

Example of the basic structural segment of a silicone.

Ceramics see their widest use in devices as components of implants, particularly orthopedic (bone replacement) implants. The ceramics used in orthopedic applications include both the manmade (or artificial) ones and the natural ceramic of bones, dentine and enamel— hydroxyapatite. All such ceramics may be either bio-inert or bio-active, based on whether they are chemically bonded to the natural bone structure or are in direct opposition to one another. Another important property of ceramics for consideration in their use in devices is whether they are porous or not. Tables 3.12 and 3.13 summarize the significant properties of the most important ceramics used in devices.

PHYSICAL PROPERTIES OF POLYMERS, ELASTOMERS, AND SILICONES As pointed out earlier, it is important to understand the physical properties of materials used in devices. Such properties are the primary concern of engineers, but the toxicologist needs to be conversant with them to understand their importance in device performance. Major physical properties and (if applicable) their ASTM designated test methods are as follows (Krause et al., 1983): Tensile Strength—the greatest longitudinal stress a substance can withstand without rupture (pounds per square inch, PSI); instrument: Monsanto Tensometer T-10 or Instron Model #1122; (ASTM D412). Abrasion Resistance—1 in.2 sample mounted in Gardner abrasion tester (Pacific Scientific) with abrasion sponge;

Figure 3.13

Road Map to Test Selections

37

run 1000 passes and report weight loss as percentage (ASTM D2486). Elongation—at breakpoint in tensile strength measurement, determine percent elongation (how much base is stretched before breaking) (ASTM D412). Modulus—during tensile strength and elongation tests, data are identified at specified percentages of elongation value (100%, 200%, 300%) and the process (PSI) measured at each elongation (ASTM D412). Tensile Set—stretch 1-inch die cut (ASTM 412c), to 300% for 3 min.; relax for 3 min.; measure difference; standard <10% (Monsanto T10 Tensometer) (ASTM F703 Sec.7.1.2). Tear Resistance—place test bar with standard partial cut in Instron Test Apparatus; stretch to propagate tear to breakage (in PSI) (ASTM D624). Wear—in vivo, measure as linear reduction in thickness at the contacting surface as determined by serial X-rays (Dickson, 1979; Dumbleton, 1981; McKellop et al., 1981, 1984; McKellop and Clarke, 1983; Seedhom et al., 1973 and Tillotson et al., 1971). Hardness (compression modulus at surface) tests are as follows: Point Penetration Test—degree of penetration of a given diameter point under a known force over a fixed time. Rockwell Hardness Tests—nonrecoverable indentation is measured after a major load is imposed for 15 sec, followed by a 10-kg load for 15 sec. The size of major load and ball diameter is used as Rockwell Hardness Durometer Test: Similar to Rockwell tests except that indenter is spring loaded and an equilibrium indentation load determined Indentation depth established the hardness. Surface characterization for pre- and post-implant evaluation should include (DePalma, 1986): Porosity Surface tension Infrared spectroscopy Scanning electron microscopy Surface charge (zeta potential) Energy dispersive X-ray analysis Surface texture analysis (stylus) Impact strength tests include pendulum, falling ball, and tensile impact tests. For pendulum tests (IZOD, Sharpey),

•

cellulosics—cotton, linen

•

polyamides (nylon) [NH(CH2)6NH-CO(CH2)4CO]

SURGILON® (ACY)

•

polyesters (polyethyleneterephthalate) [O(CH2)2OCOC6H4CO]

DACRON®

•

polyelefins (polypropylene) [CH2CH(CH3)]

SURGILENE® (ACY)

•

protein—silk

The five most commonly used nonabsorbable suture materials, with the structures and trade names of the three that are synthetic.

38

Safety Evaluation of Medical Devices

•

poly(p-diozanone) – absorbed in 6 months (rats) [O(CH2)2OCH2CO]

PDS® (Ethicon)

•

poly(glycolic acid) – absorbed in 40-90 days (rabbits) [OCH2CO2CH2CO]

DEXON® (ACY)

poly(glycoclide-co-lactide) [OCH2CO2CH2CO]90[OCH(CH3)CO2CH(CH3CO]

CICRYL® (Ethicon)

•

Figure 3.14

Three common absorbable suture materials, with their structures and trade names.

a weighted pendulum is released from fixed height and allowed to impact a fixed specimen at the bottom of the swing (notched sample). The amount of energy absorbed by the sample is then measured. Permeability of films for polymers can be determined by permeation kinetics (Berg and von Hippel, 1985): Use Franz diffusion cells (Crown Glass, Sommerville, NJ) 40-cm diameter disc of film (thickness important) Measure diffusion into liquid sink with time Can measure analytically or labelling techniques. Polymeric gel testing can be done for: Gel cohesivity—measure of ability of gel to resist flow; can measure with funnel-like orifice and measure distance of gel flow in 30 min. Gel penetration—measure of gel stiffness; a gel penetrometer measures this by recording distance that a probe falls through the gel in 5 sec. Creep resistance test characteristics include: Measurements made by placing a sample under a fixed tensile load and measuring its elongation with time. Depends on magnitude of tensile load Data often presented as elongation versus time for different stress levels.

one another at a velocity of 1 cm/s. One poise equals 100 centipoise. Viscosity in centipoise divided by the liquid density at the same temperature gives kinematic viscosity in centistokes (CS). One hundred centistokes equal one stoke. To determine kinematic viscosity, the time is measured for an exact quantity of liquid to flow by gravity through a standard capillary. Water is the primary viscosity standard with an accepted viscosity at 20°C of 0.01002 poise. Hydrocarbon liquids such as hexane are less viscous. Molasses may have a viscosity of several hundred centistokes, while for a very heavy lubrication oil the viscosity may be 100 centistokes. There are many empirical methods for measuring viscosity. A more thorough description of physical, mechanical, thermal, and chemical methods for polymer evaluation can be found in Shah (1998).

BIOLOGICALLY DERIVED MATERIALS There is an increasing use of biologically derived materials (typically from pigs and cattle) for tissue repair and prosthesis uses. While the normal concerns for artificially derived materials may be lessened for these, concerns of infection, particularly by such agents as prions and viruses, and of immunologically based tissue–tissue interactions are heightened.

Polymers Surface porosity and bacterial intrusion can be a concern if pores are 0.1 mm, and can be contaminated with microorganisms such as staphylococcus (~1 µm); streptococcus (~0.8–1 µm); pseudomonas (0.5–3 µm); and TB (0.2–0.6 µm) (Freter, 1980).

Viscosity The internal resistance to flow exhibited by a fluid, the ratio of shearing stress to rate of shear. A liquid has a viscosity of 1 poise if a force of 1 dyne/cm2 causes two parallel liquid surfaces 1 cm2 in area and 1 centimeter apart to move past

Table 3.12

FACTORS INFLUENCING TEST SELECTION Actual decisions as to what testing is to be done are based on a complex set of reasons, some of which are particular to the company involved and some of which are generally applicable. The author labels these reasons as “factors of influence,” and believes that they can be summarized as belonging to the seven categories presented in Table 3.14. The first of these factors, regulatory requirements, was extensively covered in the first chapter. The others are discussed briefly here.

Properties of Typical Ceramics Used in Devices

Single-crystal alumina Porous alumina Hydroxyapatite

Strength

Tensile modulus

Creep modulus

Fatigue life

Lubricity

Water permeability

Water absorption

High High Mod

High High Mod

High High High

High – High

Mod – –

Low High Low

Low Low Low

Chapter 3:

Table 3.13

Road Map to Test Selections

39

Ceramic Materials Used in Devices

Material Condition Source Density (g/cm3) Grain size (µm) E (tensile) (GPa) Hardness (Hv) sUFS (MPa) SUCS (MPa)

Al2O3

C

C

C

ZrO2

HP [1,2] 3.93 3–4 380 23,000 550 4,500

LTI [3] 1.7–2.2 30–40* 18–28 150–250 280–560 —

VT [3] 1.4–1.6 10–40* 24–31 150–200 70–210 —

ULTI [3] 1.5–2.2 8–15* 14–21 150–250 350–700 —

SHP [4] 6.1 <0.5 200 1300 1200 —

Abbreviations: HP: high purity; LTI: low temperature isotropic; SHP: sintered, hot isostatic pressed; ULTI: ultra low temperature isotropic; VT: vitreous (glassy); *: angstroms; —: unavailable

Perceptions It should be kept in mind that what people believe/perceive is as important as what is real. What materials are used, how a device is designed, and what testing is done are significantly influenced by current public and health care provider beliefs. Concerns about and memories of silicones, latex, toxic shock syndrome, etc., may dictate more extensive testing than regulations. Beliefs can also influence device acceptance, such as the case of IUDs (intrauterine devices) as contraceptive devices after the publicity around the Dalkon Shield.

Hazard Identification The most fundamental requirement in testing is to quickly identify (or eliminate the possibility of) any significant hazards, their services and how to eliminate or minimize them if they are present. Many of the tests used for medical devices are really designed to act as sensitive screens for hazards. They purposely maximize the potential to get a positive response (that is, they are very sensitive). Such tests share a number of common characteristics (Table 3.15) and do not establish the relevance of such findings of hazard to real life device use.

Risk Assessment The process of taking the results of toxicity and biocompatibility tests of literature findings and all other sources of information and then relating them to actual device use in the marketplace is risk assessment. The need to be able to perform a meaningful and convincing risk assessment may require the conduct of additional tests which allow for the quantification of risk (which screens usually do not). Such tests are usually focused on a single well-defined endpoint (such as mucosal irritation) as identified in a hazard identification test or screen, and have their own set of characteristics, as summarized in Table 3.15, while Table 3.16 speaks to those

Table 3.14 • • • • • • •

Factors of Influence on Safety Test Selection

Regulatory requirements Perceptions Hazard identification Risk assessment Animal welfare concerns Claims Time and economics

of specific toxicity tests. It should be pointed out at this point that all substances (even water and green apples) are toxic at some dose. The real life hazard is when the dose that harm may occur at is within the realm of likely exposure. Table 3.17 addresses the point of relative toxicity. ISO (2000) EN 14971: Application of risk management to medical devices provides specific guidance as to the performance of device risk assessments.

Claims Claims are what is said in labeling and advertising, and may be either of a positive (therapeutic or beneficial) or negative (lack of an adverse effect) nature. The positive or efficacy claims are not usually the direct concern of the toxicologist though it must be kept in mind that such claims both must be proved and can easily exceed the limits of the statutory definition of a device, turning the product into a drug or combination product. Negative claims such as “nonirritating” or “hypoallergenic” also must be proved, and are generally the responsibility of the product safety professional to provide proof for. There are special tests for such claims.

Time and Economies The final factors of influence or arbitrator of test conduct and timing are the requirements of the marketplace, the resources of the organization and the economic worth of the product. Plans for filings with regulatory agencies and for market launches are typically set before actual testing (or final stage development) is undertaken, as the need to be in the marketplace in a certain timeframe is critical. Such timing and economic issues are beyond the scope of this volume, but must be considered.

Table 3.15 Characteristics of Screens and Specific Toxicity Assays Screens • • • • • • • • •

Assay for overt toxicity Macroscopic, qualitative data Dose not related to material application Limited definition of test substance Rapid Usually single exposure Small number of replicates No internal statistical validity Minimum false negatives and maximum false positives

40

Safety Evaluation of Medical Devices

Table 3.16

but also discarding out-of-date (i.e., now known to be incorrect) information and reviewing the entire structure for connections and organization. The first step in any new literature review is to obtain as much of the following information as possible:

Characteristics of Specific Toxicity Assays

• Assay for no adverse effect level and toxic level • Quantitative data • Systematic observations on multiple end points (health behavior, nutrition, necropsy, pathology, clinical chemistry, hematology, etc.) • Specific data on strength, identity, and purity of test material • Short or long duration • Quantitative extrapolation of safely allowed • Single well defined end point • Formal internal statistical validity

PRIOR KNOWLEDGE The appropriate starting place for the safety assessment of any new chemical entity, particularly a potential new material for a medical device, is to first determine what is already known about the material and whether there are any close structural or pharmacological analogues (pharmacological analogues being agents with assumed similar pharmacological mechanisms). Such a determination requires complete access to the available literature. In using this information, one must keep in mind that there is both an initial requirement to build a data file or base, and a need to update such a store on a regular basis. Updating a data base requires not merely adding to what is already there,

Table 3.17

1. Correct chemical identity including molecular formula, Chemical Abstracts Service (CAS) Registry number, common synonyms, trade names, and a structural diagram. Gosselin et al. (1984) and Ash and Ash (1994, 1995) are excellent sources of information on existing commercial products, their components, and uses. 2. Chemical composition (if a mixture) and major impurities. 3. Production and use information. 4. Chemical and physical properties (physical state, vapor pressure, pH, solubility, chemical reactivity, etc. 5. Any structurally related chemical substances that are already on the market or in production. 6. Known or presumed biological properties. Collection of the above information is not only important for hazard assessment (high vapor pressure would indicate high inhalation potential, just as high and low pH would indicate high irritation potential), but the prior identification of all intended use and exposure patterns may provide leads to alternative information sources; for example, drugs to be used as antineoplastics or antibiotics

Classification of Chemical Hazardsa Routes of administration

Commonly used term

Single oral dose, rats LD50

Inhalation 4 hr vapor exposure mortality 1/6 – 4/6 rats (ppm)

Single application to skin of rabbits LD50

Probable lethal dose for man

Extremely toxic Highly toxic Moderately toxic Slightly toxic Practically g/kg quart or non-toxic

1 mg or less/kg 1–50 mg/kgb 50–500mg/kg 0.5–5 g/kg 5–15 g/kg

10 10–100 100–1000 1000–10,000 10,000–100,000

5 mg or less/kg 5–43 mg/kg 44–340 mg/kg 0.35–2.81 g/kg 22.6 or more g/kg

A taste, a drop, 1 grain 1 teaspoonful (4 mL) 1 ounce (30 g) 1 pint (250 g) >1 quart or >1 liter

(a)

From Deichman and Gerard (1996) and Gad and Chengelis (1998). By law, those materials with oral LD50’s of 50 mg/kg or less in rats are classified as Class B poisons and must be labeled “Poison.” Class A poisons are defined not by testing, but rather by inclusion on a regulatorily mandated list (CFR 173, section 173.326). “S 173.326 Poison A. (a) For the purpose of Parts 170–189 of this subchapter extremely dangerous poison. Class A are poisonous gases or liquids of such nature that a very small amount of the gas, or vapor of the liquid, mixed with air is dangerous to life. This class includes the following: (1) Bromactone. (2) Cyanogen. (3) Cyanogen chloride containing less than 0l9 percent water. (4) Diphosgene. (5) Ethyldichlorarsine. (6) Hydrocyanic acid (see Note 1 of this paragraph). (7) [Reserved] (8) Methyldichlorarsine. (9) [Reserved] (10) Nitrogen peroxide (tetroxide). (11) [Reserved] (12) Phosgene (diphosgene). (13) Nitrogen tetroxide-nitric oxide mixtures containing up to 33.2 percent weight nitric oxide. NOTE 1: Diluted solutions of hydrocyanic acid of not exceeding 5 percent strength are classed as poisonous articles. Class B (see S 173-343).” (b)

(b) Poisonous gases or liquids, Class A, as defined in paragraph (a) of this section, except as provided in S 173.331, must not be offered for transportation by rail express. [239 FR 18753, Dec. 29, 1964. Redesignated at 32 FR 5606, Apr. 5, 1967, and amended by Amdt. 173-94, 41 FR 16081, Apr. 15, 1976; Amdt. 173-94A, 41 FR 40883, Sept. 20, 1976.

Chapter 3:

may already have extensive toxicology data obtainable from government or private sources. A great deal of the existing toxicity information (particularly information on acute toxicity) is not available in the published or electronic literature because of concerns about the proprietary nature of this information and the widespread opinion that it does not have enough intrinsic scholarly value to merit publication. This unavailability is unfortunate, as it leads to a lot of replication of effort and expenditure of resources that could be better used elsewhere. It also means that an experienced toxicologist must use an informal search of the unpublished literature of his colleagues as a supplement to searches of the published and electronic literature. There are now numerous published texts that should be considered for use in literature-reviewing activities. An alphabetic listing of some of the more commonly used hard copy sources for safety assessment data is provided in Table 3.18. Obviously, this is not a complete listing and consists of only the general multipurpose texts that have a wider range of applicability for toxicology. Texts dealing with specialized classes of agents (e.g., disinfectants) or with specific target organ toxicity (neurotoxins and teratogens) are generally beyond the scope of this text (Wexler, 2006). Such sources can be off direct (free) internet sources (where one must beware of GIGO: garbage in, garbage out), commercial databases, and package products, to mention just the major categories. Appendix C provides addresses for major free internet sources.

MISCELLANEOUS REFERENCE SOURCES There are some excellent published information sources covering some specific classes of chemicals, for example, heavy metals, plastics, resins, or petroleum hydrocarbons. The National Academy of Science series Medical and Biologic Effects of Environment Pollutants covers 10–15 substances considered to be environmental pollutants. CRC Critical Reviews

Table 3.18

Road Map to Test Selections

41

in Toxicology is a well-known scientific journal that over the years has compiled over 20 volumes of extensive literature reviews of a wide variety of chemical substances. A photocopy of this journal’s topical index will prevent one from overlooking information that may be contained in this important source. Trade organizations such as the Fragrance Industry Manufacturers Association and the Chemical Manufacturers Association have extensive toxicology data bases from their research programs that are readily available to toxicologists of member companies. Texts that deal with specific target organ toxicity—neurotoxicity, hepatotoxicity, or hematotoxicity—often contain detailed information on a wide range of chemical structures. A published information source like the Target of Organ Toxicity series (Taylor & Francis, now halfway through a second round of revisions) is an example of a type of publication that contains important information on many industrial chemicals that may be useful either directly or by analogy. Upon discovery that the material one is evaluating may possess target organ toxicity, a cursory review of these types of texts is warranted. In the last decade, for many toxicologists the on-line literature search has changed from an occasional, sporadic activity to a semicontinuous need. Usually, nontoxicologyrelated search capabilities are already in place in many companies. Therefore, all that is needed is to expand the information source to include some of the data bases that cover the types of toxicology information one desires. However, if no capabilities exist within an organization one can approach a university, consultant, or a private contract laboratory and utilize their on-line system at a reasonable rate. It is even possible to access most of these sources from home using a personal computer. The major available on-line data bases are as follows.

National Library of Medicine The National Library of Medicine (NLM) information retrieval service contains the well-known and frequently

Published Information Sources for Safety Assessment

Title

Author, date

Report on Carcinogens, 10th Edn. Burger’s Medicinal Chemistry & Drug Discovery Carcinogenically Active Chemicals: A Reference Guide Catalog of Teratogenic Agents, 9th Edn. Chemically Induced Birth Defects Haddad and Winchester’s Clinical Management of Poisoning & Drug Overdose, 4th Edn. Clinical Toxicology Clinical Toxicology of Commercial Products, 5th Edn. Contact Dermatitis Sax’s Dangerous Properties of Industrial Materials, 10th Edn. Documentation of the Threshold Limit Values & Biological Exposure Indices, 7th Edn. Encyclopedia of Toxicology Sittig’s Handbook of Toxic and Hazardous Chemicals Medical Toxicology, 3rd Edn. The Merck Index, 14th Edn. Patty’s Industrial Hygiene and Toxicology, 5th Edn. Pharmacotherapy: A Pathophysiologic Approach, 5th Edn. Physician’s Desk Reference Registry of Toxic Effects of Chemical Substances (RTECS) Casarett and Doull’s Toxicology: The Basic Science of Poisons Toxicology of the Eye

National Toxicology Program, 2002 Abraham, 2003 Lewis, 1991 Shepard, 1998 Schardein, 2000 Haddad et al., 2007 Ford et al., 2001 Gosselin et al., 1984 Cronin, 1980 Lewis, 2000 ACGIH, 2002 Wexler, 2005 Pohanish, 2002 Dart, 2004 O’Neil et al., 2006 Bingham et al., 2001 Dipiro et al., 2005 Walsh, annual NIOSH, 2003 Klaassen, 2001 Grant et al., 1993

42

Safety Evaluation of Medical Devices

used Medline, Toxline, and Cancerlit data bases. Data bases commonly used by toxicologists for acute data in the NLM service are the following: 1. Toxline (Toxicology Information Online) is a bibliographic data base covering the pharmacological, biochemical, physiological, environmental, and toxicological effects of drugs and other chemicals. It contains approximately 1.7 million citations, most of which are complete with abstract, index terms, and CAS Registry numbers. Toxline citations have publication dates of 1981 to the present. Older information is on Toxline 65 (pre-1965 through 1980). 2. Medline (Medical Information Online) is a data base containing approximately 7 million references to biomedical journal articles published since 1966. These articles, usually with an English abstract, are from over 3000 journals. Coverage of previous years (back to 1966) is provided by back files, searchable online, that total some 3.5 million references. 3. Toxnet (Toxicology Data Network) is a computerized network of toxicologically oriented data banks. Toxnet offers a sophisticated search and retrieval package that accesses the following three subfiles: a. Hazardous Substances Data Bank (HSDB) is a scientifically reviewed and edited data bank containing toxicological information enhanced with additional data related to the environment, emergency situations, and regulatory issues. Data are derived from a variety of sources including government documents and special reports. This data base contains records for over 4100 chemical substances. b. Toxicology Data Bank (TDB) is a peer-reviewed data bank focusing on toxicological and pharmacological data, environmental and occupational information, manufacturing and use data, and chemical and physical properties. References have been extracted from a selected list of standard source documents. c. Chemical Carcinogenesis Research Information System (CCRIS) is a National Cancer Institute-sponsored data base derived from both short- and long-term bioassays on 2379 chemical substances. Studies cover carcinogenicity, mutagenicity, promotion, and cocarcinogenicity. 4. Registry of Toxic Effects of Chemical Substances (RTECS) is the NLM’s online version of the National Institute for Occupational Safety and Health’s (NIOSH) annual compilation of substances with toxic activity. The original collection of data was derived from the 1971 Toxic Substances Lists. RTECS data contains threshold limit values, aquatic toxicity ratings, air standards, National Toxicology Program carcinogenesis bioassay information, and toxicological/carcinogenic review information. The National Institute for Occupational Safety and Health is responsible for the file content in RTECS, and for providing quarterly updates to NLM: RTECS currently covers toxicity data on more than 106,000 substances.

before the chemical is to be searched. With most information retrieval systems this is a relatively straightforward procedure. Citations on a given subject may be retrieved by entering the desired free text terms as they appear in titles, key words, and abstracts of articles. The search is then initiated by entering the chemical CAS number and/or synonyms. If you are only interested in a specific target organ effect— for instance, carcinogenicity—or specific publication years, searches can be limited to a finite number of abstracts before requesting the printout. Often it is unnecessary to request a full printout (author, title, abstract). You may choose to review just the author and title listing before selecting out the abstracts of interest. In the long run, this approach may save you computer time, especially if the number of citations being searched is large. Once you have reviewed the abstracts, the last step is to request photocopies of the articles of interest. Extreme caution should be used in making any final health hazard determination based solely on an abstract or nonprimary literature source.

MONITORING PUBLISHED LITERATURE AND OTHER RESEARCH IN PROGRESS

The Merck Index is now available online for up-to-the minute access to new chemical entities.

Although there are a few other publications offering similar services, the Life Sciences edition of Current Contents is the publication most widely used by toxicologists for monitoring the published literature. Current Contents monitors over 1180 major journals and provides a weekly listing by title and author. Selecting out those journals you wish to monitor is one means of selectively monitoring the major toxicology journals. Aids available to the toxicologist for monitoring research in progress are quite variable. The National Toxicology Program’s (NTP) Annual Plan for Fiscal Year 1996 highlights all the accomplishments of the previous year and outlines the research plans for the coming year. The Annual Plan contains all projects in the president’s proposed fiscal year budget that occur within the National Cancer Institute/National Institutes of Health, National Institute of Environmental Health Sciences/National Institutes of Health, National Center for Toxicological Research/Food and Drug Administration, and National Institute for Occupational Safety and Health/Centers for Disease Control. This report includes a list of all the chemicals selected for testing in research areas that include but are not limited to mutagenicity, immunotoxicity, teratoly/reproduction, neurotoxicity, pharmacokinetics, subchronic toxicity, and chronic toxicity/carcinogenicity. Another mechanism for monitoring research in progress is by reviewing abstracts presented at the annual meetings of professional societies such as the Society of Toxicology, Teratology Society, Environmental Mutagen Society, and American College of Toxicology. These societies usually have their abstracts prepared in printed form; for example, the current Toxicologist contains over 1700 abstracts presented at the annual meeting. Copies of the titles and authors of these abstracts are usually listed in the societies/ respective journals, which, in many cases, would be reproduced and could be reviewed through Current Contents.

SEARCH PROCEDURE

NEW SOURCES

As mentioned in Chapter 1, chemical composition and identification information should already have been obtained

Scientists today are more aware than ever before of the existence of what has been called the “Information Revolutions.”

The Merck Index

Chapter 3:

At no other time in recent history has so much information become available from so many different “traditional” resources—including books, reviews, journals, and meetings—as well as personal computer-based materials such as data bases, alerting services, optical-disk-based information, and news media. The good news for toxicologists interested in the safety of chemical entities of all types is that numerous new computer-based information products are available that can be extremely useful additions to current safety and toxicology libraries. These tools enable one to save considerable time, effort, and money while evaluating the safety of chemical entities. The primary focus of this section is on the description and applications of the recent innovations of newly emerging information services based on the personal computer (PC).

KINDS OF INFORMATION The kinds of information described here are found on three types of PC media—floppy, CD-ROM, and laser disks. The products run the gamut of allowing one to assess current developments on a weekly basis, as well as carry out more traditional reviews of historical information. The general types of information one can cover include basic pharmacology, preclinical toxicology, competitive products, and clinical safety. The specific products discussed are as follows: two floppy disk-based products called Current Contents on Diskette and Focus On: Global Change; five CD-ROM products called Toxic Release Inventory, Material Safety Data Sheets, CCINFOdisc, Pollution/Toxicology, and Medline Ondisc; and a laser disk product entitled the Veterinary Pathology Slide Bank. We provide a brief synopsis of the major features of each as well as a description of their integration into a functional, PC-based Toxicology Information Center (TIC). When such a TIC is established, one will find that some unusual benefits accrue. One now has immediate and uninterrupted access to libraries of valuable and comprehensive scientific data. This access is free of “online” constraints and designed to be user friendly, with readily retrievable information available 24 hours a day, 7 days a week. The retrieved information can also usually be manipulated in electronic form, so one can use it in reports and/or store it in machine readable form as ASCII files. The minimal hardware requirements, which are certainly adequate for all items discussed here, are an IBM or IBM-compatible PC equipped with at least 640 K RAM, a single floppy disk drive, at least a 40-Mbyte hard disk drive, a CD-ROM drive, a VGA color monitor, and a printer. The basic point here is that hardware requirements are minimal and readily available. In the case of the laser disk products, a laser disk drive and high resolution (VGA) monitor are also required.

PC-Based Information Products: Floppy Disk Based We currently have ready access to a rapidly growing variety of relevant information resources. From a current awareness perspective, an excellent source of weekly information is the floppy disk-based product called Current Contents on Diskette (CCOD). Several versions are available; however, the Life Sciences version is most appropriate for this review because of its coverage, on a weekly basis, of over 1200 journals describing work in the biological sciences. One will note that

Road Map to Test Selections

43

the product has several useful features, including very quick retrieval of article citations as well as several output options (including either hard copy or electronic storage of references as well as reprint requests).

PC-Based Information Products: CD-ROM Media The gradual emergence of this technology during the past several years has recently blossomed with the introduction of several CD-ROM products that deal with safety issues surrounding the toxicology and safety of chemicals. CD-ROM media with such information can generally be characterized by two major advantages: they are relatively easy to use and are amazingly quick in retrieving data of interest.

Toxic Release Inventory (TRI) Before embarking on a discussion of products describing health, toxicology, and safety issues, it is well to be aware of a new, pilot CD-ROM version of the Environmental Protection Agency’s (EPA) 1987 Toxic Chemical Release Inventory and Hazardous Substances Fact Sheets. This TRI resource, which contains information regarding the annual inventory of hundreds of named toxic chemicals from certain facilities (since 1987), as well as the toxicological and ecological effects of chemicals, is available from the National Technical Information Service (NTIS), U.S. Department of Commerce, Springfield, Virginia 22161. The list of toxic chemicals subject to reporting was originally derived from those designed for similar purposes by the states of Maryland and New Jersey. As such, over 300 chemicals and categories are noted. (After appropriate rule making, modifications to the list can be made by the EPA.) The inventory is designed to inform the public and government officials about routine and accidental releases of toxic chemicals to the environment. The CD-ROM version of the data base can be efficiently searched with a menu-driven type of software called Search Express. It allows one to search with Boolean expressions as well as individual words and/or frequency of “hits” as a function of the number of documents retrieved on a given topic. Numerous searchable fields have been included, allowing one to retrieve information by a variety of means— for example, the compound name; the chemical registry number; the amount of material released into the air, water, or land; the location of the site of release; and the SIC code of the releasing party. One can also employ ranging methods with available numeric fields and sorting of output. It is hoped that this shared information will help to increase the awareness, concern, and action by individuals to ensure a clean and safe environment. The TRI data base is a significant contribution to that effort and the CD-ROM version is a superb medium with which to widely publicize and make accessible the findings.

Material Safety Data Sheets (MSDS) The MSDS CD-ROM is a useful resource that contains over 33,000 MSDS on chemicals submitted to the Occupational Safety and Health Administration (OSHA) by chemical manufacturers. This resource contains complete Material Safety Data Sheet (MSDS) information, as well as other important information such as the chemical formula, structure, physical properties, synonyms, registry number, and safety information. Users can easily search the CD-ROM by employing the Aldrich catalog number, Chemical Abstracts Service (CAS) number, chemical name, or molecular formula. One can also export the chemical structures to some supported software

44

Safety Evaluation of Medical Devices

for subsequent inclusion into work processing programs. The product is available from Aldrich Chemical Company, Inc., 940 West St. Paul Ave., Milwaukee, WI 54233.

Canadian Centre for Occupational Health and Safety (CCINFO) This set of four CD-ROM disks contains several valuable data bases of information that are updated on a quarterly basis: MSDS, CHEM Data, OHS Source, and OHS Data. The MSDS component currently contains over 60,000 MSDS supplied by chemical manufacturers and distributors. It also contains several other data bases [RIPP, RIPA, Pest Management Research Information System (PRIS)], one of which (PRIS) even includes information on pest management products, including their presence and allowable limits in food. A second disk in the series (CHEM Data) contains comprehensive information from the CHEMINFO, Registry of Toxic Effects of Chemical Substances (RTECS), and Chemical Evaluation Search and Retrieval System (CESARS) data bases, as well as recommendations on Transport of Dangerous Goods (TDG)/Hazardous Materials (49CFR). The third and fourth disks include Occupational Health and Safety (OHS) information. These disks contain data bases on Resource Organizations, Resource People, Case Law, Jurisprudence, Fatalities, Mining Incidents, and ADISCAN. Furthermore, information on Noise Levels, National Institute for Occupational Safety and Health (NIOSHTEC) Non-Ionizing Radiation Levels, and a Document Information Directory System is readily retrievable. These CD-ROM materials are available from the Canadian Center for Occupational Health and Safety, 250 Main Street East, Hamilton, Ontario L8N 1H6.

Pollution and Toxicology (POLTOX) This CD-ROM library also focuses our attention on environmental health and safety concerns. Scientists working in any industry or capacity that deals with toxic or potentially toxic chemicals will find it very useful. It allows one access to seven major data bases in this field in a single search through its use of “linking” features in its software. The distributors of this product have provided us with a spectrum of information dealing with toxic substances and environmental health. The collection of these data bases include five that are available exclusively from Cambridge Scientific Abstracts (CSA)—Pollution Abstracts, Toxicology Abstracts, Ecology Abstracts, Health and Safety Science Abstracts, and Aquatic Pollution and Environmental Quality. The abstracts come from journals or digests published by CSA on important issues including environmental pollution, toxicological studies of industrial chemicals, ecological impacts of biologically active chemicals, as well as health, safety, and risk management in occupational situations. The POLTOX CD-ROM contains over 200,000 records from these sources since 1981. POLTOX also contains two other useful data bases— Toxline (described earlier) and the Food Science and Technology Abstracts (FSTA) libraries. The FSTA component is a reasonably comprehensive collection of information regarding toxicological aspects of compounds found in food, including contamination, poison, and carcinogenic properties. The CD-ROM product is available from Compact Cambridge, 7200 Wisconsin Avenue, Bethesda, MD 20814.

Medline The Medline data base, which comes from the National Library of Medicine, is a superb, indispensable reference library that is particularly strong in its wide coverage of research activities in the biomedical literature. It also encompasses the areas of clinical medicine, health policy, and health care services. Each year, over 300,000 articles are reviewed and indexed into the data base. The full bibliographic citations of these articles, usually including the abstract of the published work, are available from numerous vendors in CE-ROM forma and are usually updated on a monthly basis. Information can be accessed from Medline in a variety of ways: by author, title, subject, Chemical Abstracts Service registration number, keyword, publication year, and journal title. Medline Ondisc is the CD-ROM product we employ (from Dialog Information Services, Inc., 3460 Hillview Ave., Palo Alto, CA 94304). It allows one access to the full Medline files back to 1984. Each year from that time until 1988 is covered on a single CD-ROM disk; starting in 1989, each disk covers only a six-month time period. The information is accessed through either an easily employed “menu-driven” system or a more standard on-line type of “command language.” Gower Publishing (Brookfield, VT) has published a series of “electronic handbooks” providing approved ingredient information on materials used in cosmetics, personal care additives, food additives, and pharmaceuticals. Academic Press, through its Sci-Vision branch, has just (2000) launched an ambitious service of CD ROM-based toxicity database products which are structure and substructure searchable. It is worth nothing that the CD-ROM-based system has been seamlessly integrated with (proprietary) both record-keeping and communications software so that one can optionally monitor the use of the on-line services and easily continue searching in the Dialog “on-line” environment after using the CD-ROM-based Medline library. Another very useful feature includes the storage of one’s search logic so that repetitive types of searches, over time, for example, can be done very easily.

PC-Based Information Products: Laser Disk International Veterinary Pathology Slide Bank (IVPSB) This application represents an important complementary approach toward training and awareness using laser disk technology. The IVPSB provides a quality collection of transparencies, laser videodisks, and interactive computer/ videodisk training programs. In particular, the videodisk contains over 21,000 slides from over 60 contributors representing 37 institutions from 6 countries. These slides are accessible almost instantaneously because of the tremendous storage capacity and rapid random search capabilities of the videodisk through the interactive flexibility of the computer. The information available is of particular interest to toxicologists and pathologists because the visuals illustrate examples of gross lesions of infectious diseases, regional diseases, clinical signs or external microscopy, histopathology, normal histology, cytology and hematology, and parasitology. The laser disk, a catalog of the entrees, a computer database, and selected interactive programs can be obtained from Dr. Wayne Crowell, Dept. of Veterinary Pathology, University of Georgia, Athens, GA 30602.

Chapter 3:

TYPES AND USES OF TESTS Safety assessment tests used for medical devices can generally be considered as either hazard identification/screens, or special studies uniquely designed for specific problems or types of devices. The bulk of this book will look at how each of the significant types of such tests are performed and interpreted. Table 3.19 summarizes the common varieties of available biocompatibility tests and their objectives, as well as where in the test they are considered in detail.

REASONABLE MAN The reasonable man is a concept in law which, though not universally applicable, still provides guidance as to what one can expect from those that use devices (and what, therefore, the limits are on uses for which the manufacturer of the device should be considered responsible for ensuring safety). The standard of reasonableness is obviously open to interpretation, but does provide a conceptual basis for determining what uses one must insure a device is safe for (and for one much precautionary label) and which it is not. The “test” employed in a legal sense is one of foreseeability, i.e., would a reasonable man in the defendant’s position foresee a measurable risk to the plaintiff? (Madden, 1992).

QUALIFICATIONS VERSUS PROCESS CONTROL Most of this book addresses testing from the point of view of what is done to quality a product—to get it access to the marketplace. Such testing is done, at a minimum, to meet specific regulatory requirements which one can determine by consulting the appropriate guidelines. However, biocompatibility testing does not end once a product is approved for the marketplace. Rather, some form of testing must be conducted on an ongoing basis to ensure that the lots of product

Table 3.19

Common Biocompatibility Assays

Cytotoxicity—A screen for adverse short-term biological effects. Specifically does it kill cultured maturation cells? Primary Dermal Irritation (PDI)—Evaluates potential of a single dermal exposure to cause skin irritation. Eye Irritation—Evaluates potential of a single ocular exposure to cause eye irritation. Vaginal Mucosal Irritation—Evaluates potential for a single exposure to irritate vaginal mucosa. Intracutaneous Reactivity—Evaluates to potential of a single injection of eluate to irritate or damage cutaneous tissue. Intramuscular Implant—Can be from 7 to 28 days long. Does implanted device/material damage surrounding tissue more than implanted inert “control” material? Dermal Sensitization—Evaluates the potential of a material to cause delayed contact hypersensitivity. Pyrogen/LAL—Determines if detectable levels of bacterial endotoxin/ fevers causing substances can be elated from material/device. Acute Systemic Toxicity—Determines the potential of a single dose of a material, at a predetermined level, for lethality. Carcinogenicity—Determines the potential for a material to induce and/or promote the formation of neoplasms. Mutagenicity—Determines the potential of a material to cause undesirable genetic effects. Hemolysis—Determines the potential for a product to lyse red blood cells.

Road Map to Test Selections

45

that enter the marketplace over time continue to be safe. The testing to be done to ensure such is generally specified in the device master file (DMF), but what tests are done and with what frequency is left to the judgment of the manufacturer (who is, however, charged in the GMPs with conducting an adequate program of periodic testing to ensure the continued quality and safety of the product). Such testing is usually derived from the results of qualification testing and product and (manufacturing) process validation studies (Table 3.20). Careful consideration (and statistical analysis of these and the variable that are involved in the manufacturing process generally identify which biocompatibility tests best serve to identify when the product is not as it should be, due to either the process not being in control (or there having been a series of small incremental changes which in summation have altered the process) or when changes in vendor supplied materials have occurred. A statistical analysis of the data will also clarify sampling strategies and required frequency of testing. This will lead to specification of a routine testing program for lot release, most commonly utilizing approaches shown in Table 3.21. The DMF on plant manufacturing SOPs need to specify what happens when a lot fails routine or release testing. It is sometimes wise to have a conditional two tier test scheme— an inexpensive but somewhat sensitive screening test (such as cytotoxicity) which is performed on some specified regular basis, and a second, more specific (and expensive) test which is conducted in those cases where a lot fails the screening test.

TIERS OF CONCERN: CONSUMERS, HEALTH CARE PROVIDERS, AND MANUFACTURING EMPLOYEES This book focuses primarily on the tests done to meet regulatory requirements for new product approval. Such requirements are intended to ensure the safety on the end use consumer of devices, the patients. Knowing what the intended use and claims are for a product, it is generally easy to identify what routes, duration, and extent of “exposure” or “dosing” will be. However, patients represent only the final tier of those who will be exposed to a device. There are (at least) two other tiers that we must consider. The others are the health care providers and those involved in actually producing and packaging the devices. Health care providers include nurses, doctors, laboratory technicians, pharmacists, and public health workers. Though they do not use the devices on a daily basis, they will handle and apply or administer the products. As such, they will have different routes and durations of exposure which must be considered and evaluated for safety. Dermal exposure in particular is likely to be more extensive. Likewise, those involved in manufacturing and packaging the product will have significantly different exposures. For these individuals, we must also be concerned with exposure to materials used in device construction and formulation. Here, the potential for inhalation exposure is most likely.

STERILIZATION AND CLEANLINESS It goes without saying that microbial contamination of devices must be controlled and that appropriate steps must be taken to sterilize products and materials. The subject

46

Safety Evaluation of Medical Devices

Table 3.20

Product and Process Validation

This is a series of qualification studies to demonstrate that manufacturing process controls are sufficient for preproduction quality assurance requirements and product specifications. Testing is performed to verify the effectiveness of such control and to evaluate the biological effects of processing aids added during manufacture. Environmental Control • environmental monitoring program • microorganism identification • viable and nonviable particulate analysis Manufacturing Process Control—initial qualification and ongoing control • raw material characterization (compare effects of process on characteristics determined in Phase I) • infrared analysis • cytotoxicity • physiochemical tests (USP, JP, etc.) • other materials characterization tests • bioburden testing • process water system validation • purified water monograph tests, USP • water for injection monograph tests, USP • endotoxin concentrations (LAL testing) • quality device cleaning processes • package qualifications Sterility • bioburden testing and organism identification • biological indicator studies (sport count, D-value) • sterilization cycle development • sterilization cycle validation, plan for periodic revalidation • dose determination studies (AAMI) plan for quarterly dose audits • sterility tests • EO dissipation curve studies and assessment of user exposure levels (AAMI/ISO) • package validation Finished Product Qualification—single use or reusable • physical testing for function and performance stability • chemical residues • testing for bacterial endotoxins in vitro, limulus amebocyte lysate (LAL) in vivo, rabbit pyrogen tests • biocompatibility cytotoxicity test hemocompatibility test • special material and device tests chemistry tests microbiology tests toxicology tests • nonviable particulate analysis • label claim (instructions) for reusable devices decontamination cleaning disinfection/sterilization • other product specific testing • shelf life stability qualification accelerated aging studies real time aging

is addressed later in this volume in some detail. It should also be remembered, however, that the means of sterilization (ethylene oxide, radiation, chemical sterilization or steam) may both affect device quality and, in some cases, carry their own biocompatibility concerns. Here, residuals are the issue.

Table 3.21

Routine Testing

Release testing involves what is performed routinely to satisfy GMP and ISO requirements for finished product testing prior to the release of product for distribution. In addition, Phase IV includes testing that may be incorporated into the manufacturer’s quality assurance audit program by conducting periodic raw material and finished product testing in order to document that materials and product conform to specifications. Release Testing • endotoxin concentration limulus amebocyte lysate (LAL), USP • pyrogenicity rabbit test, USP • safety test, USP infusion/transfusion assemblies • sterility testing • microbial limit test, USP • cytotoxicity, USP/ISO • materials characterization Periodic Audit Testing • endotoxin concentration limulus amebocyte lysate (LAL), USP • pyrogenicity rabbit test, USP • cytotoxicity raw materials finished products • in vitro hemolysis test for blood contract products • EO residual testing • materials characterization • physical testing • particulate testing • bioburden testing Source: HIMA, 1985.

And finally, it must be stressed that cleanliness, in the sense of exclusion of foreign matter (even seemingly innocuous things like lint and dust) is essential. If such foreign materials should gain entry to the body, they can trigger dangerous immune modulated responses (Turco and Davis, 1973). FDA has specifically considered the problem of particles in medical devices from the perspective of physiological effects and provided guidance on the issue (Marlowe, 1980).

REFERENCES Abraham, D.J. (2003). Burger's Medicinal Chemistry & Drug Discovery, 6th Edn. (6 volume set) John Wiley & Sons, New York. American Conference of Governmental Industrial Hygienists (ACGIH) (2001). Documentation of the Threshold Limit Values & Biological Exposure Indices, 7th Edn., 2002-2007 Supplements. ACGIH, Cincinnati. American Industrial Hygiene Association (AIHA) (1980). Hygienic Guide Series, Vols. I and II. AIHA. Akron. Ash, M. and Ash, I. (1994). Cosmetic and Personal Care Additives, Electronic Handbook. Gower, Brookfield, VT. Ash, M. and Ash, I. (1995). Food Additives, Electronic Handbook. Gower, Brookfield, VT. ASTM-F56. (1986). Standard specification for stainless steel sheet and strip for surgical implants. American Soc. for Testing and Materials. 13.01:7–9. Autian, J. (1980). Plastics. In: Caserett and Doull’s Toxicology: The Basic Science of Poisons. J. Doull, C.D. Klassen and M.O. Amdur (Eds.), pp. 531–556, Second Edn. McGraw-Hill, Inc., New York.

Chapter 3: Berg, O.G. and von Hippel, P.H. (1985). Diffusion-controlled macromolecular interactions. Ann. Rev. Biophys. Biophys. Chem., 14: 131–60. Billmeyer, F.W. (1971). Textbook of Polymer Science. John Wiley & Sons, NY. Bingham, E. (2001). Patty's Toxicology, 5th Edn. John Wiley & Sons, Inc., New York. Bommer, J., Waldher, R. and Ritz, E. (1983). Silicone storage disease in long-term hemodialysis patients. Contr. Nephrol., 36: 115–126. Cronin, E. (1980). Contact Dermatitis. Churchill Livingston, Edinburgh. Dart, R.C., (2004). Medical Toxicology, Third Edn., Lippincott Williams & Wilkins, Baltimore, MD. Dart, R.C., et al (Eds.) (2004). Medical Toxicology, 3rd Edn. Lippincott Williams & Wilkins. Philadelphia, PA. Deichmann, W. and Gerard, H. (1996). Toxicology of Drugs and Chemicals. Academic Press, New York. DePalma, V.A. (1986). Apparatus for zeta potential measurement of rectangular flow cells. Rev. Sci. Instrum., 51: 1390–1395. Dickson, G. (1979). Physical and chemical properties and wear. J. Dent. Res., 58: 1535–1543. Dipiro, J.T., Talbert, R.L., Yee, G.C. et al. (2005). Pharmacotherapy: A Pathophysiologic Approach, 6th Edn. McGraw-Hill, New York. Dumbleton, J.H. (1981). General considerations in friction and wear measurement. In: Tribology of Natural and Artificial Joints, pp. 110–48. Elsevier, New York. Dumitriu, S. (1994). Polymeric Biomaterials. Marcel Dekker, New York. Elbert, J.G., McKinney, P.W., Conn, Jr., J., Binder, P. and Beal, J.M. (1971). Polyglycolic acid synthetic absorbable sutures. Amer. J. Surg., 121(5): 561–565. Finkel, A.J. (1983). Hamilton and Hardy’s Industrial Toxicology, Fourth Edn. John Wright PSG Inc., Boston. Ford, M.D., Delaney, K.A., Ling, L.J., and Erickson, T. (2001). Clinical Toxicology. W.B Saunders Company, Philadelphia. Frazza, E.J. and Schmitt, E.E. (1971). A new absorbable suture. J. Biomed. Mat. Res. Sym., 1: 43–58. Freter, R. (1980). Mechanism of association of bacteria with mucosal surfaces. Ciba Found. Sump., 80: 36–55. Gad, S.C. and Chengelis, C.P. (1998). Acute Toxicology, Second Edn. Academic Press, San Diego, CA. Gosselin, R.E., Smith, R.P., Hodge, H.C (1984). Clinical Toxicology of Commercial Products, 5th Edn. Williams & Wilkins, Baltimore. Grant, Morton W., and Schuman, Joel S. (1993). Toxicology of the Eye, Fourth Edn. Charles C. Thomas, Springfield, IL. Grant, W.M., (1993). Toxicology of the Eye, 4th Edn. Charles C. Thomas, Springfield, IL. Haddad, L.M., Shannon, M.W., Borron S.W., and Burns, M.J. (2007). Haddad and Winchester's Clinical Management of Poisoning & Drug Overdose, 4th Edn. W.B Saunders Company, Philadelphia. Hardman, B. and Torkelson, A. (1986). Silicones. In: H.F. Mark et al. (Eds.), Encyclopedia of Polymer Science and Engineering, pp. 204–308. John Wiley & Sons, New York. Haslam, J., Willis, H.A. and Squirrell, D.C.M. (1972). Identification and Analysis of Plastics, Second Edn. Ififfe Books, London. Heggers, J.P., Kassovsky, N., Parsons, R.W., Robson, M.C., Pelley, R.P. and Raine, T.J. (1983). Biocompatibility of silicone implants. Ann. Plast. Surg., 11: 38–45. HIMA Report 85–1. (1985). Guidelines for the Preclinical Safety Evaluation of Materials Used in Medial Devices (T.J. Henry, Ed.), Health Industry Manufacturers Association, Washington, DC. Institute for Scientific Information (ISI) (1986). Current Contents—Life Sciences. ISI, Philadelphia. ISO (1997). ISO 10993–16 Toxicokinetic Study Design for Degradation Products and Leachables. Kambric, H.E., Muraboyoshi, S. and Nose, Y. (1986). Biomaterials in artificial organs. Chem. Engr. News (April 14): 30–48. Katz, A.R. and Turner, R.J. (1970). Evaluation of tensile and absorption properties of polyglycolic acid sutures. Surg. Gynecol. Obstet., 131: 701–16. Klaassen, C.D. (2001). Casarett and Doull's Toxicology: The Basic Science of Poisons, McGraw-Hill, New York.

Road Map to Test Selections

47

Krause, A., Lange, A. and Ezrin, M. (1983). Plastics Analysis Guide. Hansu Publishers, New York. Kroschwitz, J.I. (1990). Concise Encyclopedia of Polymer Science and Engineering. John Wiley and Sons, New York. Laohapand, T., Osman, E.M., Morley, A.R., Ward, M.K. and Kerr, D.N.S. (1982). Accumulation of silicone elastomers in regular dialysis. Proc. EDTA, 19: 143–52. Lewis, R.J. (1991). Carcinogenically Active Chemicals: A Reference Guide. Van Nostrand Reinhold, New York. Lewis, R.J. (2000). Sax’s Dangerous Properties of Industrial Materials, 10th Edn. Van Nostrand Reinhold, New York. Mackison, F., National Institute for Occupational Health and Safety/Occupational Safety and Health Administration (1981). Occupational Health Guidelines for Chemical Hazards. Department of Health and Human Services (NIOSH)/Department of Labor (OSHA) DHHS no. 81–123, Government Printing Office, Washington, DC. Madden, M.S. (1992). Toxic Torts Handbook. Lewis Publishers, Boca Raton, Florida. Marlowe, D.E. (1980). Particles in Medical Devices. U.S. Food and Drug Administration. Springfield, VA: Natl. Tech. Information Service (PB81–131625), U.S. Dept. Commerce. McGregor, R.R. (1953). Silicones in Pharmacy. Pharmacy International, January, pp. 24–26, 63. McKellop, H., Clarke, I., Markolf, K. et al. (1981). Friction and wear properties of polymer, metal, and ceramic prosthetic joint materials evaluated on a multi-channel screening device. J. Biomed. Mater. Res. 15: 619–53. McKellop, H., Hosseninian, A., and Burgoyne, K. (1984). Polyethylene wear against titanium alloy compared to stainless steel and cobalt-chromium alloys. Transactions of the Second World Congress on Biomaterials—Tenth Annual Meeting of the Society of Biomaterials. Washington, DC, April 27; May 1. McKellop, H.A. and Clarke, I.C. (1983). Evolution and evaluation of materials—screening machines and joint simulators in predicting in-vivo wear phenomena. In: P. Ducheyne and G. W. Hastings (Eds.), Functional Behavior of Orthopedic Biomaterials. CRC Press, Boca Raton, FL. McMurrer, M.C. (Ed.) (1985/1986). Plastics Compounding Redbook. HBJ Publications, pp. 14–74. Miripol, J.E. and Stern, I.J. (1977). Decreased accumulation of phthalate plasticizer during storage of blood as packed cells. Transfusion, 17: 71–2. National Institute for Occupational Safety and Health. NIOSH Criteria for a Recommended Standard for Occupational Exposure to XXX. Department of Health, Education and Welfare, Cincinnati, OH. National Institute for Occupational Safety and Health. (19XX). NIOSH Current Intelligence Bulletins. Department of Health, Education and Welfare, Cincinnati, OH. National Institute for Occupational Safety and Health (1984). Registry of Toxic Effects of Chemical Substances, Eleventh Edn., Vols. 1–3. Department of Health and Human Services DHHS No. 83–107, 1983 and RTECS Supplement DHHS 84–101, Washington, DC. National Toxicology Program (2000). Nineteenth Annual Report on Carcinogens. Department of Health and Human Services, PB 85–134633. Washington, DC. National Toxicology Program (1996). Annual Plan for Fiscal Year 1996. Department of Health and Human Services, NTP-85-0055. Government Printing Office, Washington, DC. National Toxicology Program (1999) Review of Current DHHS, DOE, and EPA Research Related to Toxicology. Department of Health and Human Services, NTP-85–056. Government Printing Office, Washington, DC. National Institute for Occupational Safety and Health (NIOSH) (2001). Registry of Toxic Effects of Chemical Substances. Washington, DC. National Toxicology Program (2005). Report on Carcinogens, 11th Edn. Department of Health and Human Services. Northup, S.J. (1989) Current problems associated with toxicology evaluation of medical device materials and future research needs, Fund. Appl. Toxicol., 13: 196–204.

48

Safety Evaluation of Medical Devices

O’Neil, M.J., Heckelman, P.E., Koch, C.B., Roman, K.J. et al. (2006). The Merck Index, 14th Edn. Merck & Co, Inc. New Jersey. Parker, C.M. (1988). Available Toxicology Information Sources and Their Use. In: S.C. Gad (Ed.), Product Safety Evaluation Handbook. Marcel Dekker, NY, pp. 23–41. Peck, C.C., Odom, D.G., Friedman, H.I., Albro, P.W., Hass, J.R., Brody, J.T. and Jess, D.A. (1979). Di-2-ethylehexyl phthalate (DEHP) and mono-1-ethylhexyl phthalate (MEHP) accumulation in whole blood and red cell concentrates. Transfusion, 19: 137–416. Petersen, M.C., Vine, J., Ashley, J.J. and Nation, R.L. (1981). Leaching of 2-(2-phdroxyethylmercapto)benzothiazole into contents of disposable syringes. J. Pharm Sci., 70: 1139–1143. Pohanish, R.P. (2002). Sittig’s Handbook of Toxic and Hazardous Chemicals, 4th Edn. Noyes Publications/William Andrew Publishing. Norwich, New York. Proctor, N.H. and Hughes, J.P. (1978). Chemical Hazards of the Workplace. J.B. Lippincott, Philadelphia. Rochow, E.G. (1951). An Introduction to the Chemistry of the Silicones. John Wiley & Sons, New York. Sasakawa, S. and Mitomi, Y. (1978). Di-2-ethylhexylphthalate (DEHP) content of blood or blood components stored in plastic bags. Vox Sang., 34: 81–86. Sax, N.I. (2000). Dangerous Properties of Industrial Materials, Tenth Edn. Van Nostrand Reinhold, New York. Schardein, J.L. (1999). Chemically Induced Birth Defects, Third Edn. Marcel Dekker, NY. Schardein, J.L. (2000). Chemically Induced Birth Defects, 3rd ed. New York: Marcel Dekker, Inc. Schmitt, E.E. and Polistina, R.A. (1967). Surgical Sutures. U.S. Patent 3,297,033, American Cyanamid. Seedhom, B., Dowson, D. and Wright, V. (1973). Wear of solid phase formed high density polyethylene in relation to the life of artificial hips and knees. Wear, 24: 35. Shah, V. (1998). Handbook of Plastics Testing Technology, John Wiley, New York. Shepard, T.H. (1998). Catalog of Teratogenic Agents, 9th Ed. Johns Hopkins University Press, Baltimore.

Sidhu, K.S., Stewart, T.M. and Netton, E.W. (1989). Information sources and support networks in toxicology. J. Amer. Coll. Toxicol., 8: 1011–1026. Steinemann, S.G. (1980). Corrosion of surgical implants—in vivo and in vitro tests. In: G.D. Winter, J.L. Leroy and K. de Groot (Eds.), Evaluation of Biomaterials, pp. 1–34. John Wiley & Sons, New York. Test method for Rockwell hardness of plastics and insulating materials. (1984). In: ASTM Annual Book of Standards. American Society of Testing Materials, Philadelphia. Procedure D-785-65. Test method for rubber property-durometer hardness (1984). In: ASTM Annual Book of Standards. American Society of Testing Materials, Philadelphia, Procedure D-2991–71. Test method for water absorption of plastics. (1984). In: ASTM Annual Book of Standards. American Society of Testing Materials, Philadelphia. Procedure D-570–81. Test methods for impact resistance of plastics and electrical insulating materials (1984). In: ASTM Annual Book of Standards. American Society of Testing Materials, Philadelphia. Procedure D-244–80. Test methods for tensile, compressive and flexural creep and creep rupture of plastics (1984). In: ASTM Annual Book of Standards. American Society of Testing Materials, Philadelphia. Procedure D-2990–77. Thomas Healthcare (2007). Physician’s Desk Reference. Thomas Healthcare, Montvale, NJ. Tillotson, E.W., Craig, R.G. and Peyton, F.A. (1971). Friction and wear of restorative dental materials. J. Dent. Res., 50: 149–154. Turco, S. and Davis, N.M. (1973). Clinical significance of particular matter: A review of the literature. Hosp. Pharm., 8:137–10. Walsh (Annual). Physicians Desk Reference. Thomson Healthcare Inc., New Jersey. Wexler, P. (Ed.) (2005). Encyclopedia of Toxicology. Elsevier Ltd., Oxford. Wexler, P. (2005). Encyclopedia of Toxicology. Academic Press, San Deigo. Wexler, P. (Ed.) (2000). Information Resources in Toxicology, Third Edn. Elsevier, NY. Wolff, M.E. (1997). Burger’s Medicinal Chemistry, Fifth Edn. John Wiley and Sons, NY.

4 Materials in Medical Device Design INTRODUCTION

METALS

Devices have been an integral part of human health care for thousands of years. There is evidence of the use of acupuncture needles in prehistoric Peruvian mummies, and surgical instrumentation used in cranial trepanations dating back thousands of years. Materials for medical devices must meet stringent requirements, including nontoxicity, sterilizability, biostability or controllable degradation, and high purity. Manufacturing processes and storage conditions must be chosen properly to prevent or minimize undesirable changes in material properties; therefore, biocompatibility assessment must be performed on the finished device as intended for use. Understanding of the interactions between living tissues and its constituents (proteins, ligands, cells) and biomaterials is rapidly evolving and providing a rational basis for device materials selection. The original concept of a biomaterial as a passive and inert nonparticipant has been expanded to include a wide range of functional and responsive materials that direct or participate in physiological processes. Modern biomaterials design is a multidisciplinary collaboration involving chemists, physicists, materials scientists, engineers, biologists, and physicians. New medical devices may take advantage of new nano-scale materials, environmentally responsive polymers, or shape memory materials. On the horizon lies a whole array of novel combination products that unite synthetic or manufactured materials with biological materials or drugs for a wide range of applications, including controlled drug delivery and regenerative medicine. The desire for less or noninvasive surgical procedures is driving evolution of new devices that can be deployed and controlled in the absence of an incision. Regulatory agencies are working feverishly to meet the need for new standard test methods and guidance to insure the safety and effectiveness of these revolutionary products. Medical devices, even combination products, contain components that fall into one of three broad categories:

A variety of metals see significant use in medical devices, though their use in patient contacting situations is largely limited to implants, with the exception of stainless steel. These materials are used primarily for orthopedic applications, e.g., total joint replacement, but are finding new applications, including cardiovascular stents and artificial spinal discs.

Stainless Steel Customarily considered the most biocompatible metal, stainless steel is by far the most common metal used in medical devices. This ferrous alloy is widely used in surgical instruments, wire sutures, needles, screws, and implant parts where great strength is needed. Steels are, by definition, alloys of iron containing 0.002% to 1.5% carbon, with stainless steel containing high percentages (∼10–25%) of chromium. This results in the formation of a passive oxide film in the presence of oxygen. Figure 4.1 displays the protective effect of passivation on steels. Corrosion resistance increases proportionally with chromium content. Other elements that are common components of stainless steel are nickel, molybdenum, and magnesium. Stainless steels are classified by microstructure into three categories, ferritic, martensitic, and austenitic, and by chemical composition into two categories, chromium type and chromium–nickel type. Ferritic stainless steels are not used as implants due to their inferior mechanical properties and inherent ferromagnetism, while use of martensitic steels is limited due to inferior corrosion resistance. Table 4.1 presents the ASTM specifications of the two grades of stainless steel commonly used for medical devices. Stainless steel rarely presents a biocompatibility problem in medical devices. In some long-term implants, however, there has been evidence of reaction of the alloy with body fluids, but even this is unusual. The recent increase in persons who have developed an allergic sensitivity to nickel (believed to have been brought about by the prevalence of nickel containing costume jewelry and the daily use of many nickel containing utensils) and the mounting evidence of nickel carcinogenicity have led to a movement to ban or restrict the use of nickel-containing alloys for specific medical applications (Sumita et al., 2004). This has propelled the development of nickel-free high nitrogen content stainless steels that meet or exceed the strength and corrosion resistance of 316L stainless steel. Cobalt-based alloys have been in use since the early 1900s as dental castings. Cobalt-based cast alloys quickly became the metal of choice for dental work over the more expensive gold that was previously used. This early cobalt alloy was a mixture of cobalt, chromium, and molybdenum, commonly known as Vitallium. Wrought cobalt alloys (cobalt, chromium, nickel, and molybdenum) with more desirable mechanical properties than the early cast alloys are used in the heavy load-bearing stems of hip implants.

• Metals are electropositive elements that are usually lustrous or shiny. They are generally good conductors of heat and electricity, and have high stiffness. Metals can be melted, fused, hammered into thin sheets, or drawn into wires for use as devices. Typical metals form salts with nonmetals, basic oxides with oxygen, and alloys with one another. • Ceramics are non-metallic, inorganic solids, including metal oxides. They are hard, brittle, heat-resistant and corrosion-resistant materials. • Polymers are composed of long or very large molecules. They are typically organic, relatively flexible, and low density. Their properties are highly temperature dependent. Polymers can be shaped by processing in solution or in the molten state. The unique properties and attributes of each of these categories of materials as they relate to their applicability for medical devices is addressed in the following chapter. 49

50

Safety Evaluation of Medical Devices

Figure 4.1

Protective action of oxide layer through metal passivation.

Stainless steel and cobalt-based alloys have elastic moduli an order of magnitude higher than that of bone. The mismatch of moduli between bone and implanted metals may result in resorption and weakening of the surrounding bony tissue. As a result, titanium has replaced stainless steel and cobalt alloys as the material of choice for many total joint replacement implants because of its lower modulus, as well as its excellent biocompatibility and enhanced corrosion resistance. Although its disadvantages include low shear strength and less than ideal wear resistance, the physical properties of titanium can be tailored to the users’ need by adjusting the percentages of alloying elements and controlling the microstructure of the alloy via processing techniques. Titanium and its alloys are widely used in a range of biomedical devices today (Long and Rack, 1998). Commercially pure titanium (made up almost entirely of titanium with trace amounts of carbon, iron, oxygen, hydrogen, and nitrogen) is used in cardiovascular stents and other soft tissue applications because of its low strength and fracture resistance as compared to alloyed titanium. A titanium alloy that is nearly ubiquitous in the medical world is Ti–6Al–4V. Originally developed for the aerospace industry, it was adopted for orthopedic implants because of its physiologically appropriate modulus and excellent corrosion resistance. Recent concerns about vanadium leaching and toxicity have led to development of novel titanium alloys specifically for biomedical use. These alloys make use of elements such as Nb, Ta, Zr, and Mo, and have lower moduli than the Ti–6Al–4V alloy. For each of the three types of titanium in use (commercially pure, Ti–6Al–4V, and medically Table 4.1

Chemical Composition of Stainless Steel Composition (%)

Element Carbon Manganese Phosphorus Sulfur Silicon Chromium Nickel Molybdenum Nitrogen Copper Iron Source: ASTM F56, 1986.

Grade 1

Grade 2

0.08 max 2.00 max 0.030 max 0.030 max 0.75 max 17.00–19.00 12.00–14.00 2.00–3.00 0.10 max 0.50 max Balance

0.030 max 2.00 max 0.030 max 0.030 max 0.74 max 17.00–19.00 12.00–14.00 2.00–3.00 0.10 max 0.50 max Balance

specific alloys) the addition of oxygen and mechanical processing results in finer microstructure grains, yielding enhanced mechanical properties and biocompatibility through improved cell adhesion (Rack and Qazi, 2006).

Nitinol Nitinol is an up and coming titanium alloy that can be used in a wide range of applications. It is a “superelastic” material that can return to its original shape after a significant deformation. This property makes it ideal for medical devices that require large recoverable deformations, such as cardiovascular and gastro-intestinal stents. Although it is a near equiatomic metal alloy of titanium and nickel, it has more than ten times the elasticity of stainless steel. As a result, Nitinol stents conform to and deform with blood vessel walls without damaging the surrounding tissue. Perhaps the most well-known characteristic of Nitinol is its temperature-induced shape memory. At low temperatures, the metal is flexible in the martensitic phase, but returns to its original shape in the higher temperature austenitic phase. The temperature at which the phase change occurs can be optimized by adjustment of the alloy composition and through heat treatment. This characteristic allows for the creation of a whole new range of biomedical tools and devices ranging from deflectable (or bendable) surgical tools to be used in arthroscopic surgery to self-expanding stents that can be inserted via a catheter (Ponsonetta et al., 2006). Despite the many potential uses for Nitinol there are still characteristics of the material that need to be further investigated. There is concern over how the surface properties of the Nitinol are affected by processing steps such as sterilization. Surface characteristics influence both biocompatibility and corrosion, specifically the leaking of nickel ions and/or the presence of oxides on the surface of the implant. While preliminary studies have been performed on how surface properties of Nitinol are affected and to what extent they are affected, more research needs to be performed before Nitinol can become ubiquitous throughout in the medical industry (Es-Souni et al., 2005).

Bioactive Metals Bioactive metals are formed by treating commonly used alloys with NaOH solutions and then heat treating them. In the case of titanium, this results in a sodium titanate gel layer that forms on the surface on the implant. This layer contains functional groups that will react with body fluid to form an

Chapter 4:

Table 4.2

Materials in Medical Device Design

51

Summary of the Physical and Mechanical Properties of Various Implant Materials in Comparison to Natural Bone

Property 3

Density (g/cm ) Elastic modulus (GPa) Compressive yield (MPa) Fracture toughness (MPA m1/2)

Natural bone

Magnesium

Ti alloy

Co–Cr alloy

Stainless Steel

Synthetic hydroxyapatite

1.8–2.1 3–20 130–180 3–6

1.74–2.0 41–45 65–100 15–40

4.4–4.5 110–117 758–117 55-115

8.3–9.2 230 450–1000 –

7.9–8.1 189–205 170–310 50–200

3.1 73–117 600 0.7

Source: Compiled from references [16-19,20]. Magnesium as an orthopedic biomaterial—From Staiger et al., 2006

integrated apatite layer that is chemically bonded with not only the surrounding bone, but also the metal deeper in the implant. Other metal oxide layers that form a functional gel layer when treated with NaOH and heat are: SiO2, TiO2, ZrO2, NbO2, and TaO2 (Kokubo et al., 2004).

Other Metals There are a variety of other metals in use in the medical and dental industries. Among these are magnesium, tantalum, and a variety of others used both as alloying elements and by themselves to address the needs of the biomedical community. Gold, palladium and platinum, for example, are plated onto the surface of stents to provide radiopacity for placement and monitoring. Silver is used as an antimicrobial.

Magnesium The use of magnesium-based biomaterials as degradable implants represents a significant departure from the traditional approaches to metallic implants. In this case, corrosive reactions resulting in the complete resorption of the metal and subsequent replacement with natural tissue are highly desirable, and a critical component in the design of the implant. Magnesium and its alloys are completely broken down by the body in a matter of weeks or months. The stiffness and yield strength of magnesium is lower than that of other biocompatible alloys by an order of magnitude, and in fact, are much closer to those of bone (see Table 4.2). Additionally, magnesium is present as a naturally occurring ion in bone and has been shown to promote osteogenesis when used as a surface ion on implants. The bioresorption and compliance match with bony tissue offer the obvious advantages of allowing bone ingrowth and reduction of stress shielding; however, if the implant degrades too quickly, complications can arise at the implant site (Staiger et al., 2006; Muller et al., 2007).

Tantalum Tantalum was initially used for dental implants in the early 1900s and has more recently been incorporated into joint replacement implants as coatings on other metals to aid in bone growth and integration. Tantalum is not only a biocompatible material but it is also bioactive, with a layer of apatite forming on its surface within several weeks of implantation. Due to its high porosity, low modulus of elasticity and relatively high frictional characteristics, it is an ideal implant material that can often bear normal physiological loads very soon after implantation. The variability of pore size can be optimized to allow not only for the in-growth of bone but also for fibrous in-growth, which allows for repair of tendon and ligament insertion sites in bone (Levine et al., 2006).

CERAMICS Ceramics are products manufactured by the action of heat on earthy raw materials (i.e., those that are neither metallic nor

organic), predominantly including silicon and its oxides, aluminum oxides, and complex materials known as silicates. In physical structure, ceramics may be polycrystalline, glass, combinations of multicrystals with glassy phases, or single crystals. The universal properties of ceramics that account for their enduring utility include: • mechanical strength in spite of brittleness • chemical durability (at both normal and elevated temperatures) against deteriorating effects of oxygen, water (liquid or vapor), acids, bases, salts and organic solvents • hardness contributing to resistance against erosion • the ability to be combined with metals and other materials to make uniquely effective devices and device parts. The forming of artificial ceramics involves particle technology, including size reduction (commination), measurement, and separation, blending and packing of particles, surface chemistry and physics, rheology (the flow of more or less plastic materials under pressure), and bonding of particles. Other sciences contribute to the understanding and control of heat treatment. Ceramics came into being because of the useful rheological properties of the clay–water system, with its capability of being shaped, whether as a paste or as a fluid suspension or as a damp, crumbly powder. Modern industrial ceramics often replace clay with organic or inorganic additives to make the raw material easier to form and to provide strength before firing (final heating). Clay-based ceramics undergo both chemical-mineralogical and physical changes during heat treatment to form new phases, including a glassy phase that often constitutes the bond between the grains in the new microstructure. Such heat treatment involves sintering in the presence of a liquid phase. The complex series of occurrences accompanying heat treatments include gas evolution, oxidation, valence changes, chemical reactions, phase transformations, melting, shrinkage, and still other chemical and physical changes. The trend with modern ceramics technology is to simplify such changes by establishing the chemical and mineral changes before forming. Maximum density is sought in the forming batch, and compaction is achieved by applying the principles of particle packing. Then heat treatment brings about sintering in the solid state with a minimum of shrinkage, gas evolution, and attendant flaws, such as warping, cracking, or the formation of unwanted pores. Many significant ceramic products, such as thermal insulation and filters for liquids and gases, require controlled porosity, both in terms of percent by volume and in pore diameters and distributions. In fact, texture compromises the prime distinction between refractories, whitewares, and electronic and glass-ceramics that may otherwise be identical in chemical and mineral composition. Ceramics see their widest use in devices as components of implants, particularly orthopedic (bone replacement) implants. The ceramics used in orthopedic applications

52

Safety Evaluation of Medical Devices

Table 4.3

Properties of Typical Ceramics Used in Devices

Single-crystal alumina Porous alumina Hydroxyapatite

Strength

Tensile modulus

Creep modulus

Fatigue life

Lubricity

Water permeability

Water absorption

High High Mod

High High Mod

High High High

High – High

Mod – –

Low High Low

Low Low Low

include both manmade (artificial) ceramics and the natural ceramic of bones, dentine and enamel—hydroxyapatite. All such ceramics may be either bioinert or bioactive, based on whether they are chemically bonded to the natural bone structure or are in direct opposition to one another. Another important consideration in the use of ceramics for devices is their porosity. Tables 4.3 and 4.4 summarize the significant properties of the most important ceramics used in devices.

b. Synthetic i. Thermoplastic polymers: nylon, polyvinyl chloride, polyethylene (linear), polystyrene, polypropylene, fluorocarbon resins, polyurethane, acrylate resins ii. Thermosetting polymers: polyethylene (cross-linked) phenolics, alkyds, polyesters c. Semisynthetic cellulosics (rayon, methylcellulose, cellulose acetate) and modified starches (starch acetate, etc.).

POLYMERS Polymers, also known as “macromolecules,” are large molecules that are formed by chemical bonding between many (poly) small molecules (mers). These bonds may be covalent or ionic in nature. Biological polymers such as proteins, polysaccharides, and DNA are the building blocks that form all living things, and play a critical role in the interaction between the body and medical devices. Synthetic polymers are not found in nature, but share many of the same characteristics of their natural counterparts. Polymers used in medical devices may be either natural (derived from plants or animals) or synthetic. Historically, most polymers used in medical devices have been plant-based natural or synthetic polymers; however, combination products that incorporate polymeric biologicals (extracellular matrix proteins, growth factors, etc.), either for improved biocompatibility or control of cellular response, are emerging as a major new class of regulated medical devices. A partial list of polymers by type includes the following (Billmeyer, 1971): 1. Inorganic: siloxane, sulfur, black phosphorus, boron– nitrogen, silicones 2. Organic a. Natural i. Polysaccharides: starch, cellulose, pectin, seaweed gums (aga, etc.), vegetable gums (arabic, etc.) ii. Polypeptides (proteins): casein, albumin, globulin, keratin, insulin, DNA (biologicals or synthetic) Table 4.4

iii. Hydrocarbons: rubber and gutta percha (polyisoprene), also called elastomers

Ceramic Materials Used in Devices

Material

Al2O3

C

C

C

ZrO2

Condition Source Density (g/cm3) Grain size (µm) E (tensile) (GPa) Hardness (Hv) σ UFS (MPa) ∑UCS (MPa)

HP [1,2] 3.93 3–4 380 23,000 550 4500

LTI [3] 1.7–2.2 30–40* 18–28 150–250 280–560 –

VT [3] 1.4–1.6 10–40* 24–31 150–200 70–210 –

ULTI [3] 1.5–2.2 8–15* 14–21 150–250 350–700 –

SHP [4] 6.1 <0.5 200 1300 1200 –

Abbreviations: HP: high purity; LTI: low temperature isotropic; SHP: sintered, hot isostatic pressed; ULTI: ultra low temperature isotropic; VT: vitreous (glassy); *: angstroms; –: unavailable.

For most devices, we are concerned only with the synthetic organic polymers. The principal class of natural polymers of concern, the elastomers, will be considered later in this chapter along with the synthetic members of this class. The chief class of inorganic polymers of concern, the silicones, will likewise be considered later in this chapter. This review will focus exclusively on polymeric materials that are not considered “biologicals” by the FDA. The reader is referred to Chapter 16 for more information on combination products. Polymers may be classified according to their main chain composition, polymerization mechanism, processibility, and bulk properties. Bulk chemical, physical and mechanical properties are a function of the properties of the chains and their arrangement within the bulk material. Polymers may be either homopolymers (characterized by a single unique monomer repeat or structural unit) or copolymers (characterized by no single unique monomer repeat or structural unit). Copolymers are often designed to incorporate or blend desirable properties of homopolymers. The polymerization mechanism for most polymers is by either condensation (or step-growth) polymerization, a chemical reaction that links monomers to yield polymers plus small molecules, or an addition (or chain growth) polymerization, in which vinyl type polymers react via attack of the C–C double bond by either free radical, anionic, or cationic chain ends. The polymerization mechanism plays a major role in determining the polymer degradation mechanism. The term “constitution” refers to the types and arrangements of main-chain and substituent atoms, the numbers and types of bonds, the composition of the chain endgroups, the presence or absence of branching or side reactions, the molecular weight and molecular weight distribution, and the mode of monomer addition. The atoms that make up the main chain play a strong role in determining the chemical properties of the bulk polymer, including solubility and wettability. The numbers and types of bonds are key determinants of the physical and mechanical properties of the bulk, while the chain endgroups influence thermal stability and degradation rate. Polymer molecular weight is strongly correlated with the processibility (viscosity of polymer melt or solution) and the mechanical properties (strength) of the bulk material. For a given polymer, an

Chapter 4:

Table 4.5

Materials in Medical Device Design

Commonly Used Homopolymers in Medical Devices

Polyethylene Polypropylene Polybutylene Polyvinylchloride

Polystyrene Polychloroprene Polyacrylates Polytetrafluoroethylene

Polysiloxanes Polysulfones Polyamides

CH2

CH

CH2

CH

CH3

Source: Autian, 1980.

CH3

CH2

CH

CH3

Figure 4.3 Structural diagram of polypropylene, a typical homopolymer, along with that of its monomeric unit (propylene).

increase in molecular weight is generally associated with increased viscosity and tensile strength. Polymer properties are also significantly affected by chain configuration and conformation. Configuration (stereochemistry, tacticity) is the spatial arrangement of substituent groups about the main chain carbon atoms which can only be changed by breaking and reforming bonds. Polymer chains that are tactic (regular and repeating arrangement of substituents groups around the main chain) can form crystalline regions, while atactic chains cannot. Conformation is the distinct spatial position adopted by atoms or groups of chain segments via rotation about single bonds. The ability of a polymer chain to make conformational changes determines the rate at which phase transitions (e.g. melting) occur and the amount of energy that a polymer can absorb in a high impact loading. Important thermal properties for polymers include the glass transition temperature, the melting temperature, and the crystallization temperature. The glass transition temperature indicates the transition between glassy or brittle and rubbery or flexible behavior in non- or semi-crystalline polymers. The melting point is the temperature at which a crystalline polymer undergoes the transition from a solid to a liquid. Polymers that are semi-crystalline exhibit both glass transition temperatures and melting points. Polymers that can be repeatedly processed and reprocessed by heating are known as thermoplastics. Examples of medical devices made from thermoplastics include sutures, medical textiles, and molded plastics. Polymers that form irreversible crosslinks upon heating are known as thermosets, and include such materials as bone cements and splints. Rubbery polymers, or elastomers, are crosslinked polymers that extend significantly and reversibly upon deformation. If the crosslinking is physical, occuring via formation of small crystalline regions amidst a matrix of noncrystalline polymer (rather than by formation of permanent covalent bonds between chains), the polymer is known as a thermoplastic elastomer. Elastomers are used in numerous

acrylonitrile

H2C

CH — C

vinyl chloride

H2C

CH — CI

butadiene

H2C

CH — CH

methyl methacrylate

H2C

C

C

CH3

O

devices, including medical examination gloves, bandages, syringe tips and sensors. Crosslinked polymers that are highly hydrophilic (water loving) are known as gels. These polymers may absorb significant amounts of water (up to 99% by weight!) and have poor mechanical properties. Typical uses for these polymers are as contact lenses and wound dressing materials. Table 4.5 lists the most commonly encountered homopolymers. Figure 4.2 provides the structures of some typical monomers, while Figures 4.3–4.7 provide structural illustrations of some homopolymers. Examples include polyvinyl acetate (Fig. 4.8), polyesters (Fig. 4.9), and polyamides (Fig. 4.10). Synthetic elastomers such as SBR synthetic rubber, made from styrene and butadiene, are also copolymers. Terpolymers, such as ABS (Fig. 4.11), are copolymers that consist of three different monomers. Linear polymers are processible into fibers, films, and molded objects. Branched polymers, which have numerous long chains projecting from the main chain, do not form strong or durable bulk materials. When polymer chains become completely chemically cross-linked (also called cured or gelled), the polymer becomes “set” and is no longer processible. This occurs when each chain is covalently bonded to another chain at a point or multiple points other than at the endgroups. A good example of this is an automobile tire, which is technically one single polymer molecule! Chemical crosslinking may be induced by heat or other forms of energy, and may involve functional groups that are already part of the polymer chain or additional chemicals that form bonds between two separate but adjacent chains. Polymers may also form cross-linked networks via physical entanglement, ionic bonding (ionomers), or by

N

CH2 O

CH3

CH2 CH CH3

Figure 4.2

53

Molecular structures of four typical isomers.

n

Figure 4.4 Structure of polypropylene, a typical homopolymer, rendered as a molecular model.

54

Safety Evaluation of Medical Devices

CH2

CH

CH2

CH

CH2

Made from: styrene

CH

CH

Figure 4.5 Structural diagram of polystyrene, a typical homopolymer, along with that of its monomeric unit (styrene).

crystallization of blocks or segments of copolymers (thermoplastic elastomers). This cross-linking is reversible, and therefore these types of polymers are processible after curing. The principle concerns with the biocompatibility of polymers are additives, residual monomers, and contaminants which are leachable in the body. As Table 4.6 shows, polymers themselves generally have very low toxicities. Partially as a reflection of their high molecular weights, true polymers themselves are not generally absorbed into the body, are not irritating, and are not sensitizers. Residual monomers, additives, and contaminants are a very different story. Each of these can be quite significant biologically. Residual monomers are the remaining individual building block units in homopolymers, copolymers, terpolymers, etc. (as diagrammed structurally in Figs. 4.2–4.11) that are not successfully incorporated into the plastic during the synthesis process. Technically, we should also include dimers, trimers, and other small-chain fragments which are left in the polymer mass but are not chemically bound to it. Many factors help determine how much residual monomer will be left in a polymer, as well as how available such residuals are to a surrounding biological matrix. And some of the

CH2 CH2

CH2 CH

CH2

n

Cl

n

Figure 4.7 Structure of polyvinyl chloride (PVC), a typical homopolymer, rendered as a molecular model along with the structure of the monomeric unit (vinyl chloride).

monomers are quite active biologically. When we test a plastic for biocompatibility, biologically available (leachable) residual monomers are a significant part of our concern. Examples of toxic monomers (and their principal toxicities) which can be found in polymers include the following: Acrylonitrile: human carcinogen (liver, brain) Vinyl chloride: human carcinogen (liver) Formaldehyde: animal carcinogen (nasal) Methylene dianline: suspect human carcinogen A wide variety of other chemical entities are specifically incorporated into plastics to achieve desired goals of structure, performance, and processing ease. Table 4.7 presents a short list of the major categories of additives. Such additives can be quite significant biologically, and may be very biologically available. A historical example is diethylhexylphthalate (DEHP), a once widely used plasticizer that was found both to be an animal carcinogen and to migrate readily from plastic bags and tubing into the blood and intravenous solutions they contained (Peck et al., 1979; Sasakawa and Mitomi, 1978). Table 4.8 shows the variety of reported concentrations of DEHP in samples of blood and blood products. The wide range of values can be attributed primarily to a difference in storage time and conditions amongst the samples. During polymer synthesis and the forming of components and products, materials such as solvents and mold release agents can contaminate the polymer product (US FDA; Petersen et al., 1981). The result of the additives and contaminants being in plastic is that a range of toxic materials may be leached from many plastics. Table 4.9 presents a short list of some of the more significant of these. For the interested reader, Kroschwitz (1990) and Dumitriu (1994) are excellent texts on polymers, particularly as they are utilized in medical devices.

ELASTOMERS

Figure 4.6 Structure of polyethylene, a typical homopolymer, with a structural diagram of its monomer, ethylene.

The classical definition of an elastomer is a member of a class of synthetic thermosetting polymeric substances possessing rubber-like qualities (such as the ability to regain shape after deformation, to be stretched to at least twice its original length, and having toughness and resistance to weathering

Chapter 4:

CH2

CH

CH2

O

CH

CH2

CH

CI

C

Materials in Medical Device Design

CH2

CH

O

CI

C

O

CH3

55

O

CH3 O

Made from: H2N

vinyl acetate

CH

H2C

vinyl chloride

OC

CH

CH3

CI

Figure 4.8 Structural diagram of vinyl acetate, a typical copolymer, along with the structures of its two constituent monomers (vinyl acetate and vinyl chloride).

and chemical attacks). For our purposes, however, the natural rubbers will also be included under the term. In fact, all thermoplastic polymers act as elastomers as temperatures above the glass transition temperature. The chemical requirements for useful elastomers include high molecular weight, flexibility of the polymer chain, and a controlled degree of polarity and structural irregularity. These features impart cohesiveness and a lack of molecular order in the undeformed state, although natural and synthetic rubbers generally take on some aspects of crystallinity at very low temperatures or when they are highly extended. The mechanical properties of cross-linked polymers are a function of the degree of cross-linking and the molecular weight between cross-links. More highly cross-linked elastomers tend to be stiffer and have higher retractive forces. Natural rubber is a natural organic substance obtained from the sap of certain trees. Rubber is produced from a wide variety of plants, especially Hevea brasiliensis (family Euphorbiaceae), a tall softwood tree native to Brazil, but also from species of Mabea, Manihot, Sapium, Ficus, and others. The milky liquid (latex) found in the inner bark and obtained by tapping (cutting) the bark is coagulated to create rubber; the oldest and simplest technique is by smoking over a fire. Plantation rubber, widely grown in Southeast Asia (particularly Malaysia), comes from trees grown from

O

(CH2)X

O

C

(CH2)y

selected seeds, with trees spaced well apart and tapping similar to the method used for wild trees, but coagulation is accomplished chemically. Most rubber used in medical devices is still of the natural variety due to economics and the better resistance of the natural product to heat. Recent concerns about latex allergies, however, have increased the move towards synthetic products. Synthetic rubbers or elastomers comprise a family of elastic products derived chiefly from petroleum and alcohol, some of which closely resemble natural rubber and some of which have quite different physical properties. Research on the chemistry of natural rubber in the 19th century led to the isolation of isoprene, which was reconverted into a rubber-like substance by distillation. This and other polymerization processes by which long chains of molecules were created attracted continued research into the early 20th century. Among the most important synthetic elastomers are: • • • • • •

C

the butadiene–styrene copolymers the various butadiene copolymers the chloroprene polymers (neoprene) the polysulfide polymers (Thiokol) cis-1,4-polyisoprene ethylene–propylene terpolymers (EPDM rubber)

O

(CH2)X

O

O

O

C

(CH2)y

O

Made from:

Figure 4.9

glycols

HO

di-acids

HOOC

(CH2)X (CH2)y

OH COOH

Structural diagram of a polyester, a typical copolymer, along with the structures of its two constituent monomers (glycols and di-acids).

56

Safety Evaluation of Medical Devices

(CH2) X

NH

NH

(CH2) y

C

C

(CH2) X

NH

NH

O

O

O

C

Made from:

Figure 4.10

diamines

H2N

di-acids

HOOC

(CH2) X

NH2

(CH2) y

COOH

Structural diagram of a polyamide, a typical copolymer, along with the structures of its two constituent monomers (diamines and di-acids).

other than carbon include the polysulfide rubbers (Thiokols) and the silicones. We will consider the silicones as a separate class of compounds. Synthetic rubbers, like natural rubbers, can be toughened by vulcanization (cross-linking) and improved and modified for special purposes by reinforcement with other materials. Elastomers are used as closures in vessels; stoppers in syringes; and in gloves, tubing, condoms, and in a host of ways in other devices. The degree of cross-linking (curing) of elastomers is particularly important in determining their biocompatibility. Natural rubbers are natural products, of course, and therefore subject to significant lot-to-lot variability in composition.

• polyurethane rubber • the isobutylene polymers (butyl rubbers) • the silicone rubbers produced by blending silicones with natural and synthetic rubbers • plasticized vinyl chloride • vinyl acetate polymers and copolymers. These can be chemically cross-linked with sulfur, peroxides, or similar agents. The term was later extended to include uncross-linked polyolefins that are thermoplastic; these are generally known as TPO rubbers. Their extension and retraction properties are notably different from those of thermosetting elastomers. Many important elastomers are copolymers in which the main molecular chain is composed of carbon atoms; they are usually based on butadiene and other organic compounds obtainable from petroleum: in Buna S, the second component is styrene, in Buna N it is acrylonitrile. Butyl rubber, used in pneumatic inner tubes, is a copolymer of isobutylene and isoprene. Elastomeric materials in which the main polymer chain contains atoms

CH2

CH

CH2

CH

Silicones Silicones (organosiloxane polymers) are a large group of siloxane polymers based on a backbone of alternating silicon and oxygen atoms with any of a variety of organic substituents attached to the silicone, such as shown in Figure 4.12 (Hardman and Torkelson, 1986; Rochow, 1951).

CH

CH2

CH2

CH

CN

y

x

z

Made from:

Figure 4.11 and styrene).

acrylonitrile

H2C

CH

CN

butadiene

H2C

CH

CH

styrene

H2C

CH

CH2

Structural diagram of ABS, a typical terpolymer, along with the structures of its three constituent monomers (acrylonitrile, butadiene,

Chapter 4:

Table 4.6

Oral Lethalities of Common Polymers

Polymer

>8 >8 >40 >20 >25 >3 >8.2 >7.5

Source: Autian, 1980.

Silicones are very widely used in devices and the pharmaceutical industry. In such uses the majority of the substituent organic groups are methyl (McGregor, 1953). Silicones have an unusual array of properties. Chief among these are thermal and oxidative stability and physical properties largely unaffected by temperature. Other important characteristics include a high degree of chemical inertness, resistance to weathering, good dielectric strength, and low surface tension. As the general formula implies, the molecular structure can vary considerably to include linear, branched, and cross-linked structures. These structural forms and the substituent (R) groups provide many combinations of useful properties that lead to a wide range of commercially important applications. Silicones include fluids, resins, and elastomers. Many derived products, including emulsions, greases, adhesives, sealants, and coatings, have been developed for a large variety of uses in the device industry. Silicones can be liquids, semi-solids (gels), or solids depending on the molecular weight and degree of polymerization. Viscosity ranges from less than 1 to more than 1 million centistokes. Polymers may be straight-chain, or cross-linked with benzoyl peroxide or other free radical initiation, with or without a catalyst. They have very low surface tensions, extreme water repellency, high lubricating properties, excellent dielectric properties and are permeable to gases and highly resistant to attack and decomposition by biological systems (Heggers et al., 1983). The combination of good water repellency and high oxygen permeability makes silicone polymers rather unique. Silicones are used as adhesives, lubricants, protective coatings, wetting agents, mold release agents, surfactants, foam stabilizers, surgical membranes and implants, gaskets, and tubing in the medical device industry. They have been used as implantable devices since the late 1950s, and in recent years (in the form of prosthetic breast implants) have become of concern for purported health effects and have acted as potent litogens (see the discussion of this aspect in Chapter 20).

Plasticizers Lubricants Antioxidants Colorants Emulsifiers Stabilizers Curing agents

Additives used in Plastics UV absorbers Blowing agents Fillers Release agents Flame and fire retardants Accelerators Antistatic agents

57

Concentration of DEHP in Blood and Blood Products

Rat LD50 (g/kg BW)

Polyethylene Polypropylene Polychloroprene latex Chlorosulfonated polyethylene Polyvinyl acetate Polyacrylonitrile Polyacrylamide Aromatic polyamides

Table 4.7

Table 4.8

Materials in Medical Device Design

Blood product Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma Platelet-rich plasma Platelet-poor plasma Leukocyte-poor plasma Fresh frozen plasma Fresh frozen plasma Fresh frozen plasma Whole blood Whole blood Whole blood Whole blood Platelet concentrate Platelet concentrate Platelet concentrate Red cell concentrate Red cell concentrate Red cell concentrate Red cell concentrate

DEHP concentration mean (µm/mL)

DEHP concentration range (µm/mL)

Year of study

44.8 54.6 38.0 – – – – 72.5 172.5 266 145 181 285 26.7 12 52.5 152.5 123.4 267.0 491 54.6 44.8 152

4.3–123.1 36.8–84.9 13.8–71.9 363–545 <110 100–275 <890 – – – 106–209 290–1230 – – 25–32 11.2–339 – – 140–620 – – – – 23.4–48.8 36.8–84.9 4.3–123.1 7.2–30.4 –

1993 1985 1985 1991 1974 1974 1974 1985 1985 1981 1973 1974 1978 1978 1973 2000 1985 1981 1972 1974 1979 1979 1985 1978 2000 1985 1993 2000 1979

Source: US FDA, Safety Assessment.

Fibers Fiber and textile-based devices are well-represented in all classes of medical devices, with broad applications including sutures, connectors, wound dressing materials, tissue reinforcement, vascular grafts, surgical gowns and drapes, and tissue scaffolds. These materials may be natural cellulosic or protein fibers, manufactured (cellulosics, polypeptides), or synthetic. Often natural and synthetic fibers are blended to take advantage of the favorable characteristics of each class. Natural cellulosic fibers are most commonly employed in applications that require high fluid absorbency,

Table 4.9

Identified Toxic Materials in Polymers

Aluminum Acrylonitrile (monomer) Arsenic Benzene Benzoic peroxide Bisphenol A Cadmium Carbon tetrachloride Dibutyl tin Epoxy curing agents Ethylene dichloride Ethylene oxide Formaldehyde Source: Autian, 1980; Northup, 1989.

Ketones and hydrocarbons Lead Mercaptobenzothiazole Methyl chloride Methylene chloride Methylene dianline Nickel PAHs on carbon black Pyrene Tin Tricresyl phosphate Triphenyl phosphate

58

Safety Evaluation of Medical Devices

Table 4.10 Five of the Most Commonly Used Nonabsorbable Suture Materials with Structures and Trade Names CH3

CH3

CH3

• polyamides (nylon) •

OSi

O

Si

OSi

CH3

n CH3

• •

CH3

Figure 4.12

•

[NH(CH2)6NH-CO(CH2)4CO] polyesters (polyethylene terephthalate) [O(CH2)2OCOC6H4CO] polyolefins (polypropylene) [CH2CH(CH3)] polyolefins (ultra high molecular weight polyethylene) [CH2CH2] protein–silk

ETHILON® (Ethicon) ETHIBOND EXCEL® (Ethicon) PROLENE® (Ethicon) DYNEEMA PURITY® (DSM Dyneema) PERMAHAND® (Ethicon)

Example of the basic structural segment of a silicone.

such as bandages or gauze, or in surgical linens. Protein fibers, such as silk, may have applications as suture materials. Recently, more exotic natural fibers like bamboo and spider silk are finding applications in medical devices. These natural fibers do not usually cause biocompatibility problems if properly prepared, cleaned, and sterilized, though environmental contamination of the fiber is a possibility. Natural polymers may be chemically removed from their original or native structures, dissolved in a solvent, and subsequently extruded as “regenerated” polymer filaments. Cellulose polymer is obtained from plants or from bacteria. Regenerated cellulose (rayon) is highly absorbent, but tends to be weak, and is used primarily in hygiene applications. Another polysaccharide polymer, chitin, is derived from the shells of crustaceans. Its more soluble derivative, chitosan, is currently being used in many medical applications including controlled drug delivery and as an antimicrobial agent. Similarly, proteins, such as silk fibroin or keratin, may be isolated from animal products (including milk or tissue) for subsequent fiber formation processes. Synthetic polymer fibers can be designed to be either durable or resorbable. Resorbable polymers are degraded in vivo, usually by hydrolysis and subsequent enzymatic digestion, and have the advantage of not needing to be removed after insertion into the body. These polymers have been commercially available for two decades as suture material and as degradable orthopedic implants (bone screws and plates), and have been investigated as tissue scaffolds and controlled drug delivery matrices. Table 4.10 presents structures of commonly used nonabsorbable suture materials, and Table 4.11 those of common resorbable suture materials. There has recently been some (seemingly unfounded) concern about the use of rayon and some other synthetic fibers due to possible contamination with diaxanes.

Hydrogels Hydrogels are cross-linked polymer gels that absorb more that 30% of their weight in water. These gels can provide a favorable environment for cell growth and tissue development, or can serve as reservoirs for drugs or biologicals. The properties of the gels are sensitive to monomer and crosslinker chemistry as well as to the degree of cross-linking. These gels may be formed by cross-linking of existing polymers or copolymers, by simultaneous polymerization and cross- linking, or by polymerization of a second polymer within an existing polymer matrix to form an interpenetrating network. In situ chemical gel formation is possible via activation of free radicals by light or heat, while physical

associations may occur due to environmental pH or ionic concentration. Degradation rates of polymer gels may be tuned by varying the ratio of degradable units within the polymer. Similarly, swelling and subsequent pore (mesh) sizes are directly related to weight between cross-links and distance between cross-links in the nonswollen polymer. “Smart” or environmentally responsive hydrogels are designed to transform in response to a specific chemical or physical cue. Changes in shape and water swelling in response to local variations in temperature or pH can be particularly useful in controlled drug delivery applications. Responsive polymer gels that swell or contract in response to a stimulus may be utilized as gates to drug reservoirs. There has been a great deal of research and patient literature on responsive hydrogel-based self-regulating insulin delivery systems that respond to increases in blood glucose levels.

Shape Memory Polymers Shape memory polymers (SMPs) are smart materials that can be tailored to many medical applications. Polyurethanebased SMPs are thermoplastic polymers that exhibit large changes in elastic modulus above and below the glass transition temperature. The polymer can be reshaped in its rubbery state (above Tg), and the shape may then be set by cooling to below Tg. Upon reheating to above Tg, the polymer will once again assume its original shape. These polymers have significant advantages over shape memory metals (Nitinol), including lower density, a wide range of transition temperatures, high shape recovery of up to 400%, excellent biocompatibility, ease of processing, and low cost. Potential applications include soft catheters, custom fitted orthopedic devices (splints/casts), artificial skin, and vascular stents. A recent innovation is polyurethane-based SMP foam

Table 4.11 Five Common Absorbable Suture Materials, with their Structures and Trade Names

• poly(p-dioxanone)—absorbed in 6 months (rats) • • • •

[O(CH2)2OCH2CO] poly(glycolic acid)—absorbed in 40–90 days (rabbits) [OCH2CO2CH2CO] Poly(glycolide-co-lactide)—absorbed in 4–8 weeks [OCH2CO2CH2CO]90[OCH(CH3)CO2CH(CH3CO] poly(glycolide-co-caprolactone)—absorbed in 91–119 days [C2H2O2]m[C6H10O2]n poly(glycolide-co-trimethylene carbonate)— absorbed in 7 months [C2H2O2]m[OCOO(CH2)3]n

PDS® (Ethicon) DEXON® (Syneture) VICRYL® (Ethicon) MONOCRYL® (Ethicon) MAXON® (Syneture)

Chapter 4:

(CHEM) that has lower density and an improved rate of deformation (Sokolowski et al., 2007).

TISSUE ENGINEERING SCAFFOLDS Replacement or regeneration of diseased or damaged tissue may make many medical devices obsolete in the near future. Over the past decade, rapid development of the techniques for the generation of functional tissue and organ formation has moved us much closer toward this goal, but there are still many challenges that must be addressed. Cellular growth, proliferation, differentiation, and function are a result of complex chemical, mechanical, and physical interactions between the cells and their environment. The relationships between these interactions and their relative contributions to cellular processes are still not completely understood, but are the subject of intense research activity. Selection of an appropriate scaffold must take into account factors including: 1. Persistence—is the scaffold to be a permanent, integrated part of the tissue, or is it desirable that the scaffold be completely replaced by tissue? 2. Function—are there functional requirements (e.g., mechanical strength, controlled drug delivery, etc.) other than supporting the growing tissue, during the period of tissue growth and formation, or after? 3. Morphology—what is the desired size and shape of the scaffold, and is it possible to form the material into that size and shape? 4. Availability—what materials are readily available and easily engineered? 5. Biocompatibility—what is the response of the cells and tissues to the material and to its breakdown products? The vast majority of tissue engineering substrates are made from resorbable polymers. Opportunities for ceramics and metals are limited due to their length of persistence and poor formability, but there are a few examples, including resorbable calcium phosphate CaP ceramics for bone tissue engineering and bioactive metals. General requirements for tissue engineering scaffolds include the ability to bind cells, some level of interconnected porosity, and mechanical properties that match those of the native tissue. Common polymers used for tissue engineering include natural protein and polysaccharide gels, resorbable synthetics, cross-linked hydrogels, and fibrous webs, or some combination of these. Ceramics may be added to polymer substrates, e.g., collagen or resorbable polyesters (PLA or PLGA), to improve their osteoconductivity. There is a wide spectrum of fabrication technologies for polymer tissue engineering scaffolds (Weigel et al., 2006). Conventional methods include electrospinning of nanofiber webs, melt foaming, solvent or thermally-induced phase separation, and microsphere sintering. Rapid prototyping or solid free-form manufacturing may be used to create very specific, predefined, or highly regular structures. Radiographic images of patient anatomy can be used to design custom implants that fit precisely into defects.

NANOMATERIALS New nanometer scale materials, or “nanomaterials,” are expected to have a major impact on medical technology in

Materials in Medical Device Design

59

wide ranging applications including sensing, surface modification, drug delivery, and tissue engineering. These materials may be divided into two broad categories, organic (carbon-based) and inorganic (Roszek et al., RIVM report).

Organic Carbon-based nanomaterials include C60, the fullerene or “buckyball,” composed of sixty spherically arranged carbon atoms, and carbon nanotubes, rolled up sheets of hexagonal carbon honeycombs. Inorganic nanomaterials include inorganic fullerene-like materials, inorganic nanotubes, nanowires, quantum dots, and dendrimers. C60 has many interesting physical, chemical, electrical, and optical properties. Modification of C60 by functionalization or doping with other groups or metal atoms can significantly alter these properties. After the discovery of C60 in 1985, several other fullerene molecules were synthesized, including C70, C76, C80, and C84. The properties of these other fullerenes are less fully characterized, but they also have many potential applications. Carbon nanotubes may be considered to be elongated fullerenes with at least one dimension smaller than 100 nm. They are found as either single-walled carbon nanotubes (SWCNTs) or concentric multiwalled carbon nanotubes (MWCNTs). Diameters are typically approximately 1.4 nm for SWCNTs, but the length is restricted only by the difficulties in synthesis. Nanotubes can be formed by a number of processes, including chemical vapor deposition, electric arc discharge, and laser ablation. These materials have amazing physical properties, including an extraordinarily high tensile modulus, and tensile strength at least ten times as high as that of steel! They are flexible, resilient, and extremely tough, making the nanotubes ideal for mechanical applications that require high strength, including composite materials for orthopedic or dental implants. Carbon nanotubes behave like metals or semiconductors, depending on their structure, and in the metallic form, conduct electricity with essentially no resistance at room temperature. Synthesis processes generally yield a mixture of the two forms. Nanotubes may find applications in displays, chemical sensors, and actuators. Assessing carbon nanotube toxicity in animal studies presents significant challenges. Since the nanotubes tend to aggregate into clumps that are too large for inhalation studies, investigators have had to rely of exposure via aerosols delivered to the lungs by intratracheal instillation (ITI) (Driscoll et al., 2000; Leong et al., 1998; Sabaitis et al., 1999). This technique does not allow investigation of the effects of the nanotubes on the upper section of the respiratory tract, and may result in an uneven distribution of nanotubes in the lung. Nevertheless, the ITI method is considered an acceptable method and appears to correlate with dust inhalation studies. These tests indicate that carbon nanotubes cause changes in the lungs that may result in impairment of pulmonary function. Dermal effects were investigated by exposing human epidermal keratinocytes (HEKs) to carbon nanotubes (Shvedova et al., 2003; Monteiro-Riviere, et al., 2005). Exposed cells demonstrated dose dependent and time dependent inflammatory responses.

Inorganic There are a number of inorganic nanomaterials, including inorganic nanotubes (see Tenne and Rao, 2004), nanowires, quantum dots, and dendrimers (Roszek et al., 2005). Of these,

60

Safety Evaluation of Medical Devices

both quantum dots and dendrimers are being investigated for medical applications.

Quantum Dots Quantum dots are nanometer scale crystals composed of semiconductor or other metals, with diameters typically between 2 and 10 nm. The dots are comprised of a semiconductor core covered by one or two layers. These nanoparticles have unique and tunable absorption and emission spectra that are dependent upon both their molecular composition and their size. The dots can be conjugated with biological molecules, e.g., proteins, for biolabelling and biosensing applications. They have been used for optical imaging cells and tissues. The wide variety of quantum dots renders it impossible to make broad generalizations regarding their toxicity. Recent toxicological studies on quantum dots have implicated both physicochemical and environmental factors in determining the toxicity of a particular type of quantum dot. These include size, charge, concentration, outer coating bioactivity (capping material and functional groups), along with oxidative, photolytic, and mechanical stability (Hardman, 2006).

Dendrimers Dendrimers are repeatedly-branched, symmetrical, synthetic macromolecules consisting of a central core, and branches that lead to end groups. Dendrimers range in size from between one and hundreds of nanometers. They are produced by specific chemical reactions that lead to controlled hierarchical synthesis, adding layer by layer of increasingly dense “generations” upon the core, until growth is no longer possible because of the lack of space. The end groups from a closed membrane-like structure, and determine dendrimer solubility. The branched structure of the dendrimer results in internal cavities in which soluble molecules (drugs, dyes, radionucleides) can become entrapped by chemical reaction of the endgroups. Hydrolytic or photolytic cleavage of the bonds between endgroups can be used to release entrapped molecules. Dendrimers are used in a wide range of medical applications, including imaging, oral and transdermal drug delivery, gene delivery, and tissue engineering. Biocompatibility of dendrimers is most strongly affected by the properties of the end-groups, but is also a function of the core. As with quantum dots, the wide array of dendrimer chemistries and morphologies require evaluation of each individual compound. It has been shown that dendrimers bearing – NH2 termini display concentrationand usually generation-dependent cytotoxicity, as well as hemolytic activity. Increased branching and biocompatible terminal groups (e.g., PEG) have been shown to reduce cytoxicity. There have been few studies of immunogenicity of dendrimers, but this work is urgently needed (Duncan and Izzo, 2005).

REFERENCES ASTM (1986). Standard specification for stainless steel sheet and strip for surgical implants. American Soc. for Testing and Materials, 13.01:7–9. Autian, J. (1980). Plastics. Caserett and Doull’s Toxicology: The Basic Science of Poisons. J. Doull, C.D. Klassen and M.O. Amdur (Eds.), 531–556, 2nd edn. Macmillan Publishing Co., Inc., New York.

Billmeyer, F.W. (1971). Textbook of Polymer Science. John Wiley & Sons, New York. Case, C.P., Langkamer, V.G., James, C., Palmer, M.R., Kemp, A.J., Heap, P.F. and Solomon, L. (1994). Widespread dissemination of metal debris from implants. Journal of Bone and Joint Surgery, British Volume 76(5): 701–712. Clayton, D.G. and Clayton, F.E. (eds). (1981). Patty’s Industrial Hygiene and Toxicology, Third Revised Edn, Vols. 2A, 2B, and 2C. John Wiley & Sons, New York. Cronin, E. (1980). Contact Dermatitis. Livingston, Edinburgh Churchill. Deichmann, W. and Gerard, H. (1996). Toxicology of Drugs and Chemicals. Academic Press, New York. DePalma, V.A. (1986). Apparatus for zeta potential measurement of rectangular flow cells. Rev. Sci. Instrum., 51: 1390–1395. Driscoll, K.E., Costa, D.L., Hatch, G., Henderson, R., Oberdorster, G., Salem, H. and Schlesinger, R.B. (2000). Intratracheal instillation as an exposure technique for the evaluation of respiratory tract toxicity: Uses and limitations. Toxicological Sciences: An Official Journal of the Society of Toxicology. 55, 1: 24–35. Duncan and Izzo (2005). Advanced Drug Delivery Reviews, 57, 15: 2215–2237. Dumitriu, S. (1994). Polymeric Biomaterials. Marcel Dekker, New York. Es-Souni, M., Es-Souni, M., Fischer-Brandies, H. (2005). Assessing the biocompatibility of NiTi shape memory alloys used for medical applications. Anal. Bioanal. Chem. 38: 557–567. Hardman, B. and Torkelson, A. (1986). Silicones. In: H.F. Mark et al. (Eds.), Encyclopedia of Polymer Science and Engineering, pp. 204–308. John Wiley & Sons, New York. Hardman, R. (2006) A toxicologic review of Quantum Dots: toxicity depends on physicochemical and environmental factors, Environ. Health Perspect. 114(2): 165–172. Heggers, J.P., Kassovsky, N., Parsons, R.W., Robson, M.C., Pelley, R.P. and Raine, T.J. (1983). Biocompatibility of silicone implants. Ann. Plast. Surg., 11: 38–45. Kokubo, T., Kim, H., Kawashita, M., Nakamura, T. (2004). Bioactive metals: preparation and properties. J. Mat. Sci.: Mat. In Med. 15: 99–177. Kroschwitz, J.I. (1990). Concise Encyclopedia of Polymer Science and Engineering. John Wiley & Sons, New York. Leong, B.K., Coombs, J.K., Sabaitis, C.P., Rop, D.A. and Aaron, C.S (1998). Quantitative morphometric analysis of pulmonary deposition of aerosol particles inhaled via intratracheal nebulization, intratracheal instillation or nose-only inhalation in rats. Journal of Applied Toxicology: JAT, 18, 2: 149–60. Levine, B.R., Sporer, S., Poggie, R.A., Della Valle, C.J., Jacobs, J.J. (2006). Experimental and clinical performance of porous tantalum. Biomaterials 27: 4671–4681. Long, M., Rack, H.J. (1998). Titanium alloys in total joint replacement—a materials science perspective. Biomaterials 19: 1621–1639. McGregor, R.R. (1953). Silicones in pharmacy. Pharmacy International, Jan, pp. 24–26, 63. Monteiro-Riviere, N.A., Nemanich, R.J., Inman, A.O., Wang, Y.Y. and Riviere, J.E. (2005). Multi-walled carbon nanotube interactions with human epidermal keratinocytes. Toxicology Letters, 155(3) (Mar 15): 377–384. Muller, W.D., Nascimento, M.L., Zeddies, M., Corsico, M., Gassa, L.M., Fernandez Lorenzo de Mele, M.A. (2007). Magnesium and its Alloys as Degradable Biomaterials. Corrosion Studies Using Potentiodynamic and EIS Electrochemical Techniques. Materials Research 10(1): 5–10. Northup, S.J. (1989) Current problems associated with toxicology evaluations of medical device materials and future research needs. Fund. Appl. Toxicol. 13: 196–204. Peck, C.C., Odom, D.G., Friedman, H.I., Albro, P.W., Hass, J.R., Brody, J.T. and Jess, D.A. (1979). Di-2-ethylehexyl phthalate (DEHP) and mono-1-ethylhexyl phthalate (MEHP) accumulation in whole blood and red cell concentrates. Transfusion, 19: 137–416. Petersen, M.C., Vine, J., Ashley, J.J. and Nation, R.L. (1981). Leaching of 2-(2-phdroxyethylmercapto)benzothiazole into contents of disposable syringes. J. Pharm Sci., 70: 1139–1143.

Chapter 4: Ponsonneta, L.,Treheux, D., Lissac, M., Jaffrezic, N., Grosgogeat, B. (2006). Review of in vitro studies on the biocompatibility of NiTi alloys. Int. J. Appl. Electromagnet. Mech. 23: 147–151. Rack, H.J., and Qazi, J.I. (2006). Titanium alloys for biomedical applications. Material Science and Engineering C, 26: 1269–1277. Rochow, E.G. (1951). Introduction to the Chemistry of the Silicones. John Wiley & Sons, New York. Roszek, B., Jong, W.H. and Geertsma, R.E. (2005). Nanotechnology in medical applications: State-of-the-art in materials and devices. Sabaitis, C.P. Leong, B.K., Rop, D.A. and Aaron, C.S. (1999). Validation of intratracheal instillation as an alternative for aerosol inhalation toxicity testing. Journal of Applied Toxicology: JAT, 19, (2) (Mar–Apr): 133–140. Sasakawa, S. and Mitomi, Y. (1978). Di-2-ethylhexylphthalate (DEHP) content of blood or blood components stored in plastic bags. Vox Sang, 34: 81–86. Shvedova, A.A., Castranova, V. Kisin, E.R., Schwegler-Berry, D., Murray, A.R., Gandelsman, V.Z., Maynard, A. and Baron, P. (2003). Exposure to carbon nanotube material: Assessment of nanotube cytotoxicity using human keratinocyte cells. Journal of

Materials in Medical Device Design

61

Toxicology and Environmental Health. Part A, 66(20) (Oct. 24): 1909–1926. Sokolowski, W., Metcalfe, A., Hayashi, S., Yahia, L., Raymond, J. (2007). Medical Applications of Shape Memory Polymer. Biomed. Mater. 2: S23–S27. Staiger, M., Pietak, A., Huadmi, J. and Dias, G. (2006). Magnesium and its alloys as orthopedic biomaterials: A review. Biomaterials, 27: 1728–1734. Sumita, M., Hanawa, T., and Teoh, S.H. (2004). Development of nitrogen-containing nickel-free austenitic stainless steels for metallic biomaterials—review. Materials Science and Engineering C, 24(6–8): 753–760. Tenne, R. and Rao, C.N. (2004). Inorganic nanotubes. Philosophical Transactions.Series A, Mathematical, Physical, and Engineering Sciences, 362 (1823) (Oct 15): 2099–2125. U.S. Food and Drug Administration (USFDA). Safety Assessment of Di92-ethylhexyl)phthalate (DEHP) Released from PVC Medical Devices. Center for Devices and Radiological Health, Rockville, MD. Weigel et al. (2006). Design and preparation of polymeric scaffolds for tissue engineering. Expert Rev. Med. Dev. 3(6): 835–851.

5 What to Test: Sampling and Sample Preparation

In Chapter 3 the question of what to test was explored from the aspect of state in the development process—that is, material, component or device. In the chapters that follow this one, the primary concern will be issues of how to actually perform tests are explained in detail. But here the vital bridge between these is addressed—how to select actually individual items for testing (sampling) and how to prepare the items selected for the testing process (sample preparation). ISO 10993-12 addresses both these issues (and the selection and use of reference materials) in summary fashion.

There are four major types of sampling methods— random, stratified, systematic and cluster (Levy and Lemoshow, 1991). Random is by far the most commonly employed method in toxicology. It stresses the fulfillment of the assumption of avoiding bias. When the entire pool of possibilities is mixed or randomized, then the members of the group are selected in the order they are drawn from the pool. Stratified sampling is performed by first dividing the entire pool into subsets or strata, then doing randomized sampling from each strata. This method is employed when the total pool contains subsets which are distinctly different but in which each subset contains similar members. An example is a large batch of a powdered plastic resin which it is desired to determine the nature of the particle size distribution. Larger pieces or particles are on the top, while progressively smaller particles have settled lower in the container and at the very bottom, the material has been packed and compressed into aggregates. To determine a timely representative answer, proportionately sized subsets from each layer or strata should be selected, mixed and randomly sampled. This method is used most commonly in material studies. In systematic sampling, a sample is taken at set intervals (such as every fifth container of reagent or taking a sample of water from a fixed sample point in a flowing stream every hour). This is most commonly employed in quality assurance or (in the clinical chemistry laboratory) in quality control. In cluster sampling, the pool is already divided into numerous separate groups (such as cases of product), and we select small sets of groups (such as several boxes of product from the selected cases) then select a few members from each set. What one gets then is a cluster of measures. Again, this is a method most commonly used in quality control or in environmental studies when the effort and expense of physically collecting a small group of units is significant. In classical toxicology studies sampling arises in a practical sense in a limited number of situations. The most common of these are as follows:

SAMPLING Sampling is simply the act of selecting the appropriate items or parts to test so that what is tested (either the product, material, or an extract of either of these) is representative of what is to actually be marketed (i.e., final product as shipped) or used in production. For an approved product, of course, the sampling will be part of an ongoing quality assurance process. Sampling, the selection of which and how many individual data points will be collected, whether in the form of selecting which to select from an assembly line or to remove a portion of a lot of plastic for testing, is an essential step upon which all other efforts towards a good experiment or study are based, and is essential for quality control, and it is a statistically driven process. There are three assumptions about sampling which are common to most of the statistical analysis techniques that are used in safety assessment. These are that the sample is collected without bias, that each member of a sample is collected independently of the others and that members of a sample are collected with replacements. Precluding bias, both intentional and unintentional, means that at the time of selection of a sample to measure, each portion of the population from which that selection is to be made has an equal chance of being selected. Ways of precluding bias are discussed elsewhere (Gad, 2006). Independence means that the selection of any portion of the sample is not affected by and does not affect the selection or measurement of any other portion. Finally, sampling with replacement means that in theory, after each portion is selected and measured, it is returned to the total sample pool and thus has the opportunity to be selected again. This is a corollary of the assumption of independence. Violation of this assumption (which is almost always the case in toxicology and all the life sciences) does not have serious consequences if the total pool from which samples are sufficiently large (say 20 or greater) that the chance of reselecting that portion is small anyway.

1. Selecting a subset of animals or test systems from a study to make some measurement (which either destroys or stresses the measured system, or is expensive) at an interval during a study. This may include such cases as doing interim necropsies in a chronic study or collecting and analyzing blood samples from some animals during a subchronic study. 2. Analyzing inhalation chamber atmospheres to characterize aerosol distributions with a new generation system. 3. Analyzing diet in which test material has been incorporated. 63

64

Safety Evaluation of Medical Devices

4. Performing quality control on an analytical chemistry operation by having duplicate analyses performed on some materials. 5. Selecting data to audit for quality assurance and control purposes (Ryan, 1989; CDRH, 1992).

If the result is not critical, a simple sign test will work. For the sign test, we first determine the middle value in the numbers being checked for randomness. We then go through a list of numbers assigned to each group, scoring each as a “+” (greater than our middle number) or “−” (less than our middle number). The number of pluses and minuses in each group should be approximately equal.

RANDOMIZATION As pointed out above, randomization is an essential step in avoiding bias in sampling. Randomization is the act of assigning a number of items (plates of bacteria or test animals, for example) to groups in such a manner that there is an equal chance for any one item to end up in any one group. This is a control against any possible bias in assignment of subjects to test groups. A variation on this is censored randomization, which insures that the groups are equivalent in some aspect after the assignment process is complete. The most common example of a censored randomization is one in which it is insured that the body weights of test animals in each group are not significantly different from those in the other groups. This is done by analyzing group weights both for homogeneity of variance and by analysis of variance after animal assignment, then rerandomizing if there is a significant difference at some nominal level, such as p ≤ 0.10. The process is repeated until there is no difference. There are several methods for actually performing the randomization process. The three most commonly used are card assignment, use of a random number table, and use of a computerized algorithm. For the card-based method, individual identification numbers for items (plates or animals, for example) are placed one at a time in succession into piles corresponding to the required test groups. The results are a random group assignment. The random number table method requires only that one have unique numbers assigned to test subjects and access to a random number table. One simply sets up a table with a column for each group to which subjects are to be assigned. We start from the head of any one column for each group to which subjects are to be assigned. We start from the head of any one column of numbers in the random table (each time the table is used, a new starting point should be utilized). If our test subjects number less than 100, we utilize only the last two digits in each random number in the table. If they number more than 99 but less than 1000, we use only the last three digits. To generate group assignments, we read down a column, one number at a time. As we come across digits which correspond to a subject number, we assign that subject to a group (enter its identifying number in a column) proceeding to assign subjects to groups from left to right filling one row at a time. After a number is assigned to an animal, any duplication of its unique number is ignored. We use as many successive columns of random numbers as we may need to complete the process. The third (and now most common) method is to use a random number generator that is built into a calculator or computer program. Procedures for generating these are generally documented in user manuals. One is also occasionally required to evaluate whether a series of numbers (such as an assignment of animals to test groups) is random. This requires the use of a randomization test, of which there are a large variety. The chi-square test can be used to evaluate the goodness-of-fit to a random assignment.

SAMPLE PREPARATION How samples are prepared for testing once selected is as critical as what samples are tested. With devices, much of sample preparation centers around the derivation of extracts for use in tests where inclusion of an intact solid device is either inappropriate or not physically positive. The manufacturer should choose a biologically relevant solvent system that will yield a quantity of extract sufficient to perform tests for biocompatibility. The observed biological response results from the combination of the concentration of the substances that reach the target cells and the intrinsic activity of the substances upon these cells. No single simple extraction procedure would simulate the effect of exposure to the physiological environment. For example, serum contains electrolytes, a variety of fats, and nitrogenous compounds. However, the extraction solution should be chemically and biologically uncomplicated so that the extraction solution itself does not interfere with subsequent tests. The extraction solution should include an appropriate combination of polar and nonpolar solvents, and the extraction procedure should occur in a static condition as well as under agitation. The ratio of product to extraction solution and the time for the extraction process should reflect the quantities and dwell times occurring in the anticipated use of the device. At least two solvents, polar and nonpolar, are used to obtain soluble extracts for biocompatibility tests preferable at an elevated temperature. Polar solvent can be water, saline solution, and/or water or saline solution with alcohol. Nonpolar solvent can be cottonseed oil (CSO), sesame oil, and/or polyethylene glycol (DEG) solution. The most commonly employed solvents for extraction are characterized in Table 5.1. Culture medium may also be used for cell based test systems such as in cytotoxicity testing. Extracting conditions shall attempt to exaggerate the clinical use conditions so as to define the potential toxicological hazard without causing significant changes, such as fusion or melting of the material pieces, or alter the chemical structure. If a device component melts or loses shape during extraction, the resulting extractant solution cannot be used in a valid test. The results derived from tests where the conditions of extraction were exaggerated need to be viewed in light of these exaggerations. Judgment needs to be used in interpreting the results as to their appropriateness to the actual use conditions and device potential toxicity. The concentration of any endogenous or extraneous substances in the extract, and hence the amount exposed to the test cells, depends on the interfacial area, condition of the sample surface, the extraction volume, pH, chemical solubility, osmolarity, agitation, temperature, and other factors. These conditions should each be carefully considered. It should also be remembered that for solid polymer and elastomer components, unfinished areas (such as are exposed

Chapter 5:

Table 5.1

What to Test: Sampling and Sample Preparation

65

Commonly Utilized Extractants and Vehicles

Common Name: Cottonseed Oil (CSO) Chemical Name: NA Molecular Weight: NA Formula: Mixture of natural products; glycerides of palmitic, olive and linoleic acids Density: 0.915–0.921 g/mL Volatility: Low Solubility/Miscibility: Soluble in ether, benzene, chloroform and DMSO. Slightly soluble in ethanol. Biological Considerations: Orally, serves as energy source (and therefore can alter food consumption and/or body weight). Prolonged oral administration has been associated with enhanced carcinogenesis. Chemical Compatibility/Stability Considerations: Thickens upon prolonged exposure to air. Available in USP grade. Uses (Routes): In extractions and as a vehicle for oral, dermal, vaginal, rectal and subcutaneous administration. Common Name: DMSO/Dimethyl Sulfoxide Chemical Name: Sulfinylbis [methane]; CAS #67-68-5 Molecular Weight: 78.13 Formula: C2H6OS Density: 1.100 g/mL at 20ºC Volatility: Medium Solubility/Miscibility: Soluble in water, ethanol, acetone, ether, oils Biological Considerations: Oral LD50 (rats) = 17.9 mL/kg. Repeated dermal exposure can defat skin. Repeated oral exposure can produce corneal opacities. Not cytotoxic to cells in primary culture at less than 0.05% (V/V). Intraperitoneal LD50 (mice) = 11.6 mL/kg. Chemical Compatibility/Stability Considerations: Very hydroscopic liquid. Combustible. Uses (Routes): All, as a carrier at up to 5% to enhance absorption. Common Name: Ethanol; ETOH Chemical Name: Ethyl alcohol; CAS #64-17-5 Molecular Weight: 46.07 Formula: C2H5OH Density: 0.789 g/mL Volatility: High, but declines when part of mixture with water. Solubility/Miscibility: Miscible with water, acetone and most other vehicles. Biological Considerations: Orally, will produce transient neurobehavioral intoxication. Oral LD50(rats) = 13.0 mL/kg. Intravenous LD50 (Mice) = 5.1 mL/kg. Chemical Compatibility/Stability Considerations: Flammable colorless liquid available USP grade. Uses (Routes): Extraction solvent vehicle for dermal and oral, though can be used in lower concentrations for most other routes. Volume of oral instillation should be limited to 5 mL/kg. Common Name: Polyethylene Glycol (PEG) Chemical Name: NA Molecular Weight: 400 (approximate average, range 380–420) Formula: H(OCH2CH2)nOH Density: 1.128 g/mL Volatility: Very low Solubility/Miscibility: Highly soluble in water. Soluble in alcohol and many organic solvents. Biological Considerations: Employed as water soluble emulsifying/dispersing agents. Oral LD50 (mice) = 23.7 mL/kg. Oral LD50 (rats) = 30 mL/kg. Chemical Compatibility/Stability Considerations: Do not hydrolyze or deteriorate on storage and will not support mold growth. Clear, viscous liquid. Uses (Routes): As extraction solvent for oral administration as a vehicle full strength or mixed with water. Total dosage of PEG-400 should not exceed 5–10 mL. Common Name: Saline Chemical Name: Physiological saline; isotonic salt solution Molecular Weight: 18.02 Formula: 0l9% NaCl in water (weight to volume) Density: As water Volatility: Low Solubility/Miscibility: As water Biological Considerations: No limitations—preferable to water in parenteral applications. Chemical Compatibility/Stability Considerations: None Uses (Routes): Extraction solvent all except perocular. Source: Gad and Chengelis, 1998; Lewis, 1993.

66

Safety Evaluation of Medical Devices

if an elastomeric closure or stopper piece is cut) are likely to have more leachable materials than are present in a more fully cured or finished surface. Use of a lipophilic (such as CSO) and a hydrophilic (generally saline) solvent system simulates likely physiologic extraction conditions in use. The addition of ethanol and PEG provides a fair representation of potential extraction conditions when the device is in extended contact with a drug or therapeutic solution (Autian, 1977). General points or guidance for extraction include (AAMI, 1997): 1. The extraction shall be performed in sterile, chemically inert containers by using aseptic techniques. 2. The extraction time and temperature are dependent on the physicochemical characteristics of the material and extraction vehicle. Recommended conditions are: (a) not less than 24 hours at 37ºC (current expedation is for 72 hours); (b) 72 hours at 50ºC; (c) 24 hours at 70ºC; (d) 1 hour at 121ºC. 3. Extraction conditions should simulate as closely as possible the conditions under which the device will normally be used. Therefore item (a) gives the preferred conditions for extraction and generally the other conditions are when a device is intended for steam or thermal sterilization or resterilization (such as with surgical instruments). 4. The recommended conditions should be applied according to the device characteristics and specific conditions of use. 5. Extraction procedures using culture medium with serum can only be used under the conditions specified in (a) above (i.e., not less than 24 hours at 37ºC). It should be noted that such media has both lipophillic and hydrophilic component. 6. When agitation is considered to be appropriate, the method should be specified and reported. 7. When appropriate, cut the material into small pieces before extraction. For polymers, 10 mm% 50 mm pieces have been used. Molded elastomer closures should be tested intact and not cut prior to extraction. 8. The ratio between the surface area of the material and the volume of extraction vehicle shall be no more than 6 cm2/mL. The surface area shall be calculated on the basis of the overall sample dimensions, not taking into account surface irregularity and porosity. However, the actual surface characteristics should be considered in the interpretation of the test results. If the surface area is indeterminate, then 0.1–0.2 g/mL shall be used. 9. Liquid extracts shall, if possible, be used immediately after preparation. 10. If an extract is stored, then the stability of the extract under the conditions of storage should be verified with appropriate methods. 11. If the extract is filtered, centrifuged, or processed by other methods prior to being applied to the cells, this must be included in the final report. 12. For use in direct contract tests (such as implantation studies, materials which have various shapes, sizes or physical states (i.e., liquid or solid) may be tested without modification in the cytotoxicity assays. 13. The preferred sample of a solid specimen should have at least one flat surface. Adjustments shall be made for other shapes and physical states. 14. The sterility of the test specimen shall conform to the requirements in the USP.

15. Test materials from sterilized devices which are normally supplied non-sterile but are sterilized before use shall be sterilized by the method recommended by the manufacturer and handled aseptically throughout the extraction and test procedure. 16. The effect of sterilization methods or agents on the device should be considered in defining the preparation of the test material prior to use in the test system. 17. Test materials from devices not required to be sterile in use shall be used as supplied and handled aseptically throughout the extraction and test procedure. 18. Liquids shall be tested by either: a) direct deposition; or b) deposition on to a biologically inert absorbent matrix (filter discs have been found to be suitable). 19. If appropriate, materials classed as super-absorbent shall be wetted with culture medium prior to testing. The USP (1996) provides specific guidance for use in preparing extraction solutions for use in biological reactivity tests. These are as follows. Apparatus—for the tests includes the following. Autoclave—Use an autoclave capable of maintaining a temperature of 121 ± 2.0º, equipped with a thermometer, a pressure gauge, a vent cock, a rack adequate to accommodate the test containers above the water level, and a water cooling system that will allow for cooling of the test containers to about, but not below, 20º immediately following the heating cycle. Oven—Use an oven, preferably a forced-circulation model, that will maintain operating temperatures of 50º or 70º within ± 2º. Extraction containers—Use only containers, such as ampuls or screw-cap culture test tubes, of Type I glass. If used, culture test tubes are closed with screw caps having suitable elastomeric liners. The exposed surface of the elastomeric liner is completely protected with an inert solid disk 0.05–0.075 mm in thickness. A suitable disk may be fabricated from a polytetrafluoroethylene (polytef) resin. Preparation of apparatus—Cleanse all glassware thoroughly with chromic acid cleansing mixture, or if necessary with hot nitric acid, followed by prolonged rinsing with water. Clean cutting utensils by an appropriate method (e.g., successive cleaning with acetone and methylene chloride) prior to use in subdividing a specimen. Clean all other equipment by thorough scrubbing with a suitable detergent and prolonged rinsing with water. Render containers and equipment used for extraction, and in transfer and administration of test material, sterile and dry by a suitable process. [If ethylene oxide is used as the sterilizing agent, allow adequate time for completing degassing.]

Procedure Preparation of Sample Both the Systemic Injection Test and the Intracutaneous Test may be performed using the same extract, if desired, or separate extracts may be made for each test. Select and subdivide into portions a Sample of the size indicated in Table 5.2. Note that ISO 10993 (Part 12) guidance is equivalent but stated differently, as presented in Table 5.3.

Chapter 5:

Table 5.2

What to Test: Sampling and Sample Preparation

67

Surface Area of Specimen to be Useda

Form of material

Thickness

Film or sheet

<0.5 mm 0.5–1 mm <0.5 mm (wall) 0.5 to 1 mm (wall) >1 mm

Equivalent of 120 cm total surface area (both sides combined) Equivalent of 60 cm2 total surface area (both sides combined) Length (in cm) = 60 cm2/(sum of ID and OD circumferences) Length (in cm) = 60 cm2/(sum of ID and OD circumferences) Equivalent of 60 cm2 total surface area (all exposed surfaces combined)

Strips of about 5 × 0.3 cm

>1 mm

Equivalent of 25 cm2 total surface area (all exposed surfaces combined)

Do not subdivideb

Tubing Slabs, tubing, and molded items Elastomers

Amount of Sample for each 20 mL of extracting medium 2

Subdivided into

Sections of about 5 × 0.3 cm Pieces up to about 5 × 0.3 cm

a

When surface area cannot be determined due to the configuration of the specimen, use 0.1 g of elastomer or 0.2 g of plastic or other polymers for every 1 mL of extracting fluid. b Molded elastomeric closures are tested intact.

Remove particulate matter, such as lint and free particles, by treating each subdivided Sample or Negative Control as follows: place the Sample into a clean, glassstoppered, 100-mL graduated cylinder of Type I glass, and add about 70 mL of Water for Injection. Agitate for about 30 sec, and drain off the water, repeat this step, and dry those pieces prepared for the extraction with Vegetable Oil in an oven at a temperature not exceeding 50º. [Note—Do not clean the Sample with a dry or wet cloth or by rinsing or washing with an organic solvent, surfactant, etc.].

Preparation of Extracts Place a properly prepared Sample to be tested in an extraction container, and add 20 mL of the appropriate extracting medium. Repeat these directions for each extracting medium required for testing. Also prepare one 20-mL blank of each medium for parallel injections and comparisons. Extract by heating in an autoclave at 121º for 60 minutes, in an oven at 70º for 24 hours, or at 50º for 72 hours. Allow adequate time for the liquid within the container to reach the extraction temperature. [Note—The extraction conditions should not in any instance cause physical changes such as fusion or melting of the Sample pieces, which result in a decrease in the available surface area. A slight adherence of the pieces can be tolerated. Always add the cleaned pieces individually to the extracting medium. If culture tubes are used for autoclave extractions with Vegetable Oil, seal screw caps adequately with pressure-sensitive tape.]

Table 5.3

Extraction Ratiosa

Form/Material

Thickness

Surface area/Volume

Non-absorbent Absorbents and hydrocolloidsb Film, sheet or tubing wallc Slabs, tubing and molded items Elastomersd Indeterminate surface area

Not applicable Not applicable

6 cm2/mL

a

<0.5 mm 0.5 to 1 mm > 1 mm > 1 mm Not applicable

0.1g (1 mL + absorption capacity) 6 cm2/mL 3 cm2/mL 3 cm2/mL 1.25 cm2/mL 0.2 g sample/mL or 0.1 g elastomer/mL

ISO 10993—Part 12: Sample Preparation and Reference Materials. Based on a technique developed by N.J. Stark. c May be subdivided into strips or sections. d Do not subdivide: cut edges have different extraction properties than outer surfaces. b

Cool to about room temperature but not below 20º, shake vigorously for several minutes, and decant each extract immediately, using aseptic precautions, into a dry, sterile vessel. Store the extracts at a temperature between 20º and 30º, and do not use for tests after 24 hours. Of importance are the contact of the extracting medium with the available surface area of the plastic and the time and temperature during extraction, the proper cooling, agitation, and decanting process, and the aseptic handling and storage of the extracts following extraction. One should be aware that close reading requirements under MHW Notification 99 is required. Otherwise, exaggerated extraction conditions may be inappropriately and erroneously employed (such as in the case of biologically derived materials such as collagen).

REFERENCE MATERIALS In nearly every biocompatibility test, reference materials are used to serve as experimental controls. Negative controls, in the form of blanks, are used in most biological evaluations where test article extracts are prepared. The use of these blanks provides the basis for a comparison of the effects of the test material extract with a validated negative test result. Japanese (MHW) guidelines consistently refer to these as Standard Reference Materials (SRMs). A number of materials have been used extensively in biological testing as negative or positive controls. High density polyethylene, obtained from the U.S. Pharmacopoeia, is a standard negative control. The nonreactive plastic can be implanted into living tissue and the results compared with those for a test material that has been similarly implanted. Likewise, a polyvinyl chloride formulation containing organotin additives serves well as a positive control.

CONCLUSIONS ISO 10993-12: “Sample Preparation and Reference Materials” clearly indicates that it is preferable to evaluate medical devices in their final product form. The reasoning is simple—the biological testing must incorporate everything involved in making the device. Obviously, the constituent materials must be safe for patient contact; equally important to device biocompatibility are the processes and materials used during manufacturing. For most devices, the use of fluid extracts of the test materials prepared in a fashion to mimic

68

Safety Evaluation of Medical Devices

or exaggerate the expected clinical conditions is the most appropriate technique for determining the potential effects of chemical leachables. Extraction fluid selection, extraction conditions, and material-to-extractant ratios are all outlined in the standard. The selection and use of appropriate experimental controls also is important in evaluating device materials for safety and is also covered in ISO 10993-12.

REFERENCES AAMI (1997). Biological Evaluation of Medical Devices. AAMI, Washington, DC. Autian, J. (1977). Toxicological evaluation of biomaterials: Primary acute toxicity screening program. Artif. Org., 1: 53–60.

CDRH (1992). Regulatory Requirements for Medical Devices. FDA, Washington, DC. Gad, S.C. and Chengelis, C.P. (1998). Acute Toxicology Testing, 2nd Edn. Academic Press, San Diego, CA. Gad, S.C. (2006). Statistics and Experimental Design for Toxicologists, 4th Edn. CRC Press, Boca Raton, FL. ISO (1993). 10993—Part 12 Sample Preparation and Reference Materials. Levy, P.S. and Lemoshow, S. (1991). Sampling of Populations: Methods and Applications. John Wiley & Sons, New York. Lewis, R.J. (1993). Hawley’s Condensed Chemical Dictionary. Van Nostrand Reinhold Co., New York. MHW (1999). Guidelines for Basic Biological Tests of Medical Materials and Devices, Notification 99, MHW (Japan). Ryan, T.P. (1989). Statistical Methods for Quality Improvement. John Wiley & Sons, New York. USP (1996). United States Pharmacopoeia XXIII. United States Pharmacopoeia, Philadelphia, PA.

6 Cytotoxicity Testing

The cell culture including cytotoxicity methods is a long established screening method for device or biomaterial biocompatibility screening and may be a fair predictor of biocompatibility when used together with other appropriate tests (Wilsnack, 1976; Gad, 2000). Several highly specialized cell culture methods are available to evaluate the biocompatibility of the raw materials used in the manufacturing of the device or auditing the production of the manufacturing process. Such cytotoxicity offers several advantages: ● ● ● ●

● ●

result of exaggerated pharmacological responses. Many adverse events are the consequence of disturbance of normal physiology and do not result in cell death. This is one reason that cytotoxicity assays on their own cannot provide a full assurance of safety. At the same time, the intact organism has available extensive defensive mechanisims not available to the cell in cultures. Cytotoxicity, the causing of cell death, is often the consequence of exposure to a harmful chemical, but the number of cells which must be killed before the function of a tissue or organism is noticeably impaired is highly variable. Some cell types, notably the epithelia including the liver, have the ability to regenerate in response to insult while others, most notably neurons, have little if any such capability. Some organs, such as the liver, lung, and kidney, have a substantial reserve capacity in excess of normal requirements and normal body function can be maintained in the presence of marked impairment. Cytotoxicity assays measure loss of some cellular or intercellular structure and/or functions, including cell death. They are generally simple to perform, reproducible, and have a clearly defined endpoint. However, specificity in the prediction of end points in vivo standard (such as lethality, irritation, and implantation) tests is extremely variable for a variety of reasons, including the fact that the assay systems are continually exposed, whereas in vivo there are biological protective measures in operation. Some assays may not be universally capable of detecting all chemical classes of irritants because of the endpoint used. There are several factors to consider when selecting a cell line for these tests. These include the following:

it is simple, rather inexpensive, and easy to perform; it allows testing of a biomaterial on human tissue; it is sensitive to toxic material; it is easy to manipulate and allow more than one end-point investigation; it can be used to construct a dose–response curve; and, it can give quick and quantitative results and allows direct access or direct observation or measurement.

Despite these advantages, cytotoxicity use is currently limited to screening the biomaterials and finished medical devices. Therefore, cytotoxicity results should be used for biocompatibility in conjunction with other tests. The objective of cell cytotoxicity testing is to screen the biocompatibility of the polymer and elastomer portions of medical devices or medical device components using mammalian cell cultures. Cytotoxicity is a useful method for screening material, and frequently serves as a quality control mechanism for audit or batch testing programs. It is a basic part of all device biocompatibility evaluation (AAMI, 1997; CDRH, 1992; ISO, 1993). It is one of the oldest assays designed specifically to screen plastics for toxicity (Rosenbluth et al., 1965). Given the extreme sensitivity of this test, materials found to be cytotoxic must be assessed along with the results of in vivo studies and others to evaluate the relevant risk to human health. Unlike the other studies utilized in biocompatibility testing, cytotoxicity is not a pass/fail test in the eyes of regulators. Failure in cytotoxicity is generally grounds for performing a confirmatory test such as an implantation or intracutaneous reactivity (Barile, 1994).

● ● ● ●

● ●

● ●

BACKGROUND

● ●

Historically, cytotoxicity was originally adapted in the early 1950s for the use of screening plastics used in pharmaceutical containers. The concern was for potential toxins leaking from the containers into the drugs back before our current analytical chemistry methods were available to serve the purpose. As such the plastics were called “closures.” The great majority of toxic compounds are chemically stable and produce their characteristic effects by interference with biochemical or physiological homeostatic mechanisms. This means that a full understanding of the pharmacology of extractable toxics is essential. In the case of drugs, it has been estimated that some 80% of the adverse reactions are the

●

precise nature of the study; availability of the cell or tissue; relevance to intended use; need or acceptability of the use of primary culture, early passage cultures, or established cell line; need to use differentiated cell or cell line; requirement for metabolic activation and the compatibility with S9 or other metabolizing systems; easily identifiable markers or endpoints; ease of handling and storage; absence of special growth requirements; published information on the use; and experience.

A variety of cell lines have been used including corneal epithelial cells, lung fibroblasts, Chinese hamster ovary cells, canine renal cells, HeLa cells, and microorganisms. Some cell lines used with success by FRAME (the Fund for the Replacement of Animals in Medical Experiments) are BCL-D1 (a human embryo lung finite-lived fibroblast-like cell), CHO (Chinese hamster ovary), RL4 (rat liver), L929 (fibroblastic mouse subcutaneous), HeLa (human tumor cell line), and BHK (fibroblastic baby hamster kidney cell). The popular human cell line WI38 has been replaced with many 69

70

Safety Evaluation of Medical Devices

new cell lines or explant cultures of human origin because WI38 is no longer available in a quantity to meet the demand. The Japanese recommend the V79 cell line. Differentiated cells are used to evaluate the effects materials may have on specific tissues. Differentiated cells are generally non-fibroblastic cells, which are different from transformed and fibroblastic cells such as L-cells used in ISO cytotoxicity test methods. Differentiated cells have organ specific or tissue specific functions and have specific biological end-points or measurable characteristics. Liver cells, which are differentiated cells, have all or some liver functions. Most cells in culture are fibroblasts (sometimes unintentionally). Primary cells that are taken directly from an animal often are difficult to establish in culture and become fibroblasts losing the normal functions of growing differentiated cells. Numerous conditions have to be optimized for obtaining good growth of differentiated cells. Most cultured cells have a fibroblastic appearance, although they may not be true fibroblasts. For example, cells grown under nonoptimum conditions can temporarily take an appearance of fibroblasts. The fibroblasts in culture can take over cultures because they grow readily on plastic surfaces. Success in growing differentiated cells has been partially due to techniques that have been developed to remove and limit the growth of fibroblasts to allow other cells to grow. The properties of the cell cultures usually depend on the cultivation conditions, and normal cells can grow in culture only for a limited number of generations. To perform the test in differentiated cells is important for at least two reasons. First, the tissue-type specific features of differentiated cells may modulate the effects of chemicals on the fundamental properties of cells. Second, it is important to determine the effects of chemicals on specific cell functions or responses. Culture systems for growing epithelial, liver, or embryonic cells have been developed only recently. The number of available differentiated cells for biocompatibility testing is currently small, but there is significant development in this area. A very large number of endpoints for cytotoxicity assays have been described, some of which are given in Table 6.1. The possible cytotoxicity endpoints are (Caldwell, 1993; Northup, 1986, 1987, 1992; Gad, 2000): ●

● ● ● ● ● ● ● ● ● ● ●

microscopic examination of cell morphology, membrane integrity, and fragility; cell population and density; cell adhesiveness; cytopathic effect; total protein content; rate of growth; rate of protein synthesis; total DNA content; rate of DNA synthesis; colony forming efficiency; tryphan blue uptake and other dye uptake; and biochemical assays of enzymes.

Much of the initial work on the range of cytotoxicity assays that have been developed (and their endpoint measurement methods) was done with the goal of developing an in vitro alternative to the rabbit eye irritation test. A few of the more commonly used methods are briefly described below. In closing, it should be noted that a cytotoxicity screen is a fundamental and essential part of other in vitro toxicity tests, including all of the genotoxicity assays.

Table 6.1 Examples of Endpoint Measurements used in Cytotoxicity Assays Method

Basis/Endpoint

Reference

Microphysiometry

Metabolic rate measurement Membrane damage

Bruner et al. (1991)

General cytotoxicity

Borenfreund and Puerner (1984, 1985, 1987) Rohde (1992) Sina et al. (1992) Riddell et al. (1986) Rohde (1992) North-Root et al. (1982)

Uridine uptake inhibition Neutral Red uptake Neutral Red release Leucine incorporation Total protein Fluorescein leakage Colony forming efficiency MTT dye reduction Crystal violet staining Alkaline phosphatase release Intracellular ATP

Cell membrane injury General cytotoxicity General cytotoxicity Cell membrane injury Lethal cytotoxicity

Shopsis and Sathe (1984)

Mitochondrial damage Sina et al. (1992) Lethal cytotoxicity Itagaki et al. (1991) Membrane injury Scaife (1985) General metabolic toxicity

Kemp et al. (1985)

CRYSTAL VIOLET STAINING Itagaki et al. (1991) employed a simple technique using cultured HeLa S3 cells or SIRC cells (an established line of rabbit corneal cells) in the presence of serial dilutions of the test material. After the incubation period, crystal violet was used to stain residual viable cells. The IC50 was calculated, i.e., the concentration of test material inhibiting growth of cells by 50%. Using various surfactants they found good correlations between the IC50 and the maximum in vivo eye irritation scores for the materials tested.

SILICONE MICROPHYSIOMETER This is a light-addressable potential sensor device which can be used to indirectly measure the rate of production of acidic metabolite from cells placed in a biosensor flow chamber. The endpoint calculated is the MRD50, i.e., the concentration of test material required to reduce the metabolic rate by 50% (Bruner et al., 1991). Mouse fibroblasts have been used as the test cell. Bagley et al. (1992) found that the MRD50 for a variety of materials correlated well with the maximum average score for in vivo eye irritation.

MICROTOX TEST This test utilizes changes in luminescence from Photobacterium phosphoreum, which is generated through a process linked to respiration by NAADH and flavin mononucleotide (Bulich, et al., 1990). Light output is measured photometrically before and after the addition of the test substance and an EC50 value calculated, i.e., 50% reduction in light emission. Bagley et al. (1992) found that, in general, test substances with the highest in vivo irritation gave the lowest EC50 values.

NEUTRAL RED UPTAKE ASSAY In this procedure cells, usually mouse fibroblasts or Chinese hamster ovary cells, are exposed to the test material and then

Chapter 6:

to neutral red. Retention of neutral red indicates cell viability. Bagley et al. (1992) found that, in general, the concentration of test material required to reduce neutral red uptake decreased as the in vivo irritant potential of the test material increased. Attention needs to be paid to the technical aspects of incubation, as emphasized by Blein et al. (1991), who found that correlation with materials with an extreme pH were underestimated because of the buffering effect of the culture medium and that volatile materials were also underestimated, probably because of loss of material. Several inter-test comparisons have been undertaken. Sina et al. (1992) compared leucine incorporation, MTT dye reduction and neutral red uptake in corneal epithelial cells and Chinese hamster lung fibroblasts. None of the endpoint target cell combinations accurately predicted in vivo eye irritation in this series but the MTT dye reduction method gave the best overall correlation. Cytotoxicity assays assess the effect of test material on a particular aspect of cell or intercellular function, or of lethal cytotoxicity. They, therefore, give a measure of potential to cause cell and tissue injury and as such may be used as a screen for predicting local tissue injury, including eye injury. The appropriate choice of the test cell and endpoint indicator for certain chemical classes may give reasonable prediction of the potential for eye irritation. The choice of the screening cytotoxicity assay, or assays, should in part be determined by past experience, the likely mechanism for an irritant response, and the chemistry of the material tested. In vitro assays such as cytotoxicity serve to assess acute toxic effects, and in far larger organisms are most effective in predicting local tissue irritation effects. Such experiments using cultured cells utilize a variety of methods based on either a fresh isolate from fragments of tissue or cell suspension (primary cell culture) which grow to confluence and then age and die, or single cell clones (continuous cell culture) which have an indefinite capacity to grow and replicate. The continuous cell lines have the advantage of being consistent, reliable, and reproducible. They act as a standard with a documented history, they have fewer biological variables and may be tuned to particular toxicity concerns by using a variety of tissues and species with a range of doses and exposure periods. As a result these methods can be very efficient in screening and are often more sensitive than acute toxicity tests in animals. Early cell culture methods were markedly subjective screening to merely estimate the numbers of living or dead cells, but now morphological analysis by electron microscopy reveals a spectrum of microcellular changes; and cell function tests measure biochemical parameters indicating the nature of cell stress. Although many modifications have been made, cell culture tests are of four main types: gel diffusion, direct contact, extract dilution, and cell function tests. Gel diffusion uses agar or agarose to cover a cell monolayer. A sample of the material or extract is placed on top of the gel providing a concentration gradient of diffusibles. Agarose allows a faster diffusion of uncharged molecules and is as sensitive as the rabbit intramuscular implantation test. Direct contact of the test material on to a culture layer is more sensitive than the rabbit intramuscular implantation test but care must be taken to avoid physical damage to the cells by pressure or movement of the sample. Extracts may be serially diluted in the nutrient media and provide a quantitative comparison with reference extracts. Inevitably, the correlation with animal tests will depend on the nature of the eluants.

Cytotoxicity Testing

71

Cell function tests are a very precise way of registering cellular response to any insult. In particular, inhibition of cell growth can be measure with considerable sensitivity. With increasing complexity of test methodology, the results may be less reproducible; and increasing sensitivity may not assist the prediction of risk to humans as the impact of a material on the body systems may be much less intense than in the culture plate. As discussed above, several approaches to using tissue culture methods are available for testing of biomaterials. These are divided into two major groups: One tests the toxicity of a soluble extract of the material, and another tests the toxicity by the direct contact of cells with the material or components of the device. Examples of cytotoxicity test methods using extracts include: ● ● ●

Fluid Medium Tissue Culture Assay; Inhibition of Cell Growth Assay; and Cloning Efficiency Assay.

Fluid Medium Tissue Culture evaluates the cellular damage caused by the test extract on a confluent monolayer culture. The text extract is incorporated into the culture medium which is usually double strength (2×) Minimum Essential Medium supplemented with serum and other essential nutrients at the maintenance level. The toxic effect on the monolayer, such as cell lysis and microscopic observation of cell morphology changes, is usually checked after 24 and 48 hours. Cell lysis can be scored by direct microscopic observation or with the use of radiolabels or tryphan blue dye uptake. The inhibition of cell growth is a more informative test requiring more time and skill. Distilled water extract is incorporated into the tissue culture medium and inoculated with the cells in the tissue culture tubes. After 72 hours, the extent of cell growth is determined by total protein assay, such as the Lowery photometric method, on the removed cells from the individual tubes. The Cloning Efficiency Assay is more informative, sensitive, and quantitative and requires even more skill. The Cloning Efficiency Assay’s procedure and endpoint is similar, but is more accurate, sensitive, and direct than Cell Growth Inhibition or Fluid Medium methods. The Cloning Efficiency assay normally uses a Chinese Hamster Ovary cell line and a single-cell cloning technique to estimate the toxic insult induced in cloning efficiency. The number of cells in The initial inoculum is considerably less than those plated to get confluent cultures (approximately 1 × 105 or less cells in a 100 × 15 mm Petri dish, compared to 1 × 106 or more cells). The initial inocula are incubated for approximately 24 hours before these growing cells are treated with the tissue culture medium containing the test extract. The cytotoxic effect of the extract is determined by measuring the ability of the treated cells to form colonies during seven subsequent days of incubation. The cloning efficiency of the treated cultures is compared to that of the control. The agar overlay method can be used to evaluate the toxicity of the extracts, but it is primarily used for the direct contact cytotoxicity tests of the solid test sample. Several tests are available to test cytotoxicity by direct contact. These include: ●

●

ASTM F813 Practice for Direct Contact Cell Culture Evaluation of Materials for Medical Devices; ASTM F895 Test Method for Agar Diffusion Cell Culture Screening for Cytotoxicity;

72 ●

●

●

●

Safety Evaluation of Medical Devices

ASTM F1027 Standard Practice for Assessment of Tissue and Cell Compatibility of Orificial Prosthetic Materials and Devices; NIH Publication No. 85-2185 Guidelines for Blood– Material Interactions; HIMA Report: Guidelines for the Preclinical Safety Evaluation of Materials Used in Medical Devices; and others including many device specific toxicity guidance documents on toxicity testing.

In addition, the Agar Overlay Tissue Culture method and Fluid Medium Tissue culture method can be used for direct contact cytotoxicity testing. In the Fluid Medium method, the test material or device is placed directly on the growing monolayer cell surface. In the Agar Overlay method, the solid test sample is placed on or in the agar layer containing the vital stain, such as neutral red over the growing monolayer of cells. The response is evaluated grossly and microscopically and graded according to the zone index, the size of the cytopathic area, the lysis index, and percent of cell lysis. Proper cytotoxicity testing should include at least one test with extract and one direct contact test if feasible. In addition, differentiated cells are used to evaluate the effects materials may have on specific tissues. Differentiated cells are generally non-fibroblastic cells, which are different from transformed and fibroblastic cell lines such as L-cells used in ISO cytotoxicity test methods. Differentiated cells have organ specific or tissue specific functions and have specific biological end-points or measurable characteristics. Liver cells, which are differentiated cells, have all or some liver functions. Most cells in culture are fibroblasts. Primary cells that are taken directly from an animal often are difficult to establish in culture and become fibroblasts losing the normal functions of growing differentiated cells. Numerous conditions have to be optimized for obtaining good growth of differentiated cells. Most cultured cells have a fibroblastic appearance, although they may not be true fibroblasts. For example, cells grown under non-optimum conditions can temporarily take an appearance of fibroblasts. The fibroblasts in culture can take over cultures because they grow readily on plastic surfaces. The recent success in growing differentiated cells was partially due to techniques that have been developed to remove and limit the growth of fibroblasts to allow other cells to grow. The properties of the cell cultures usually depend on the cultivation conditions, and normal cells can grow in culture only for a limited number of generations. The test in differentiated cells is important for at least two reasons. First, the tissue-type specific features of differentiated cells may modulate the effects of chemicals on the fundamental properties of cells. Second, it is important to determine the effects of chemicals on specific cell functions or responses. Culture systems for growing epithelial, liver, or embryonic cells have been developed only recently. The number of available differentiated cells for biocompatibility testing is currently small, but significant developments in this area are coming. Information in this area can be found The three specific tests prescribed by ISO (and USP) are presented below.

Table 6.2

Reactivity Grades for Agar Diffusion Test

Grade

Reactivity

Description of reactivity zone

0 1 2 3 4

None Slight Mild Moderate Severe

No detectable zone around or under specimen Zone limited to area under specimen Zone extends less than 0.5 cm beyond specimen Zone extends 0.5 to 1.0 cm beyond the specimen Zone extends greater than 1.0 cm beyond specimen but does not involve entire dish

leachable chemicals from the polymeric specimens. Extracts of materials that are to be tested are applied to a piece of filter paper.

Sample Preparation Use extracts, prepared as directed or use portions of the test specimens having flat surfaces not less than 100 mm2 in surface area.

Procedure Prepare the monolayers in 60-mm diameter plates using 7 mL of Cell Culture Preparation. Aspirate the culture medium from the monolayers, and replace it with serum-supplemented culture medium containing not more than 2% of agar. Place the flat surfaces of Sample Preparation, USP Negative Control Plastic RS (to provide a Negative Control), and either USP Positive Bioreaction Extract RS or USP Positive Bioreaction Solid RS (to provide a Positive Control) in duplicate cultures in contact with the solidified agar surface. Incubate all cultures for not less than 24 hours at 37 × 1º, preferably in a humidified incubator containing 5 ± 1% of carbon dioxide. Examine each culture around each Sample, Negative Control, and Positive Control, under a microscope, using cytochemical stains, if desired.

Interpretation of Results The biological reactivity (cellular degeneration and malformation) is described and rated on a scale of 0 to 4 (Table 6.2). Measure the responses obtained from the Negative Control and the Positive Control. The test system is suitable if the observed response corresponds to the labeled biological reactivity grade of the relevant Reference Standard. Measure the response obtained from the Sample Preparation. The Sample meets the requirements of the test if none of the cell culture exposed in the Sample show greater than a mild reactivity (Grade 2). Repeat the test if the suitability of the system is not confirmed.

DIRECT CONTACT TEST This test is designed for materials in a variety of shapes. The procedure allows for simultaneous extraction and testing of leachable chemicals from the specimen with a serum-supplemented medium. The procedure is not appropriate for very low- or high-density materials that could cause mechanical damage to the cells.

Sample Preparation AGAR DIFFUSION TEST This test is designed for elastomeric closures in a variety of shapes. The agar layer acts as a cushion to protect the cells from mechanical damage while allowing the diffusion of

Use portions of the test specimen having flat surfaces not less than 100 mm2 in surface area.

Procedure Prepare the monolayers in 35-mm diameter plates using 2 mL of cell suspension. Aspirate the culture medium from

Chapter 6:

Table 6.3

Reactivity Grades for Direct Contact Test and for Elution Test

Grade

Reactivity

Conditions of all cultures

0 1

None Slight

2

Mild

Discrete intracytoplasmic granules; no cell lysis More than 20% of the cells are round, loosely attached, and without intracytoplasmic granules; occasional lysed cells are present More than 50% of the cells are round and devoid of intracytoplasmic granules; extensive cell lysis and empty areas between cells Greater than 70% of the cell layers contain rounded cells and/or are lysed Nearly complete destruction of the cell layers

3

Moderate

4

Severe

the cultures, and replace it with 0.8 mL of fresh culture medium. Place a single Sample Preparation, USP Negative Control Plastic RS (to provide a Negative Control), and USP Positive Bioreaction Solid RS (to provide a Positive Control) in each of duplicate cultures. Incubate all cultures for not less than 24 hours at 37 ± 1º in a humidified incubator preferably containing 5 ± 1% of carbon dioxide. Examine each culture around each Sample, Negative Control, and Positive Control, under a microscope, using cytochemical stains if desired.

Interpretation of Results Proceed as directed for Interpretation of Results under Agar Diffusion Test using Table 6.3. The Sample meets the requirements of the test if none of the cultures treated with the Sample shows greater than a mild reactivity (Grade 2). Repeat the test if the suitability of the system is not confirmed.

ELUTION TEST This test is designed for the evaluation of extracts of polymeric materials. The procedure allows for extraction of the specimens at physiological or non-physiological temperatures for varying time intervals. It is appropriate for high-density materials and for dose–response evaluations.

Sample Preparation Prepare as directed in Preparation of Extracts, using ether Sodium Chloride Injection (0.9% NcCl) or serum-free mammalian cell culture media as Extraction Solvents. If the size of the Sample cannot be readily measured, a mass of not less than 0.1 g of elastomeric material or 0.2 g of plastic or polymeric material per mL of extraction medium may be used. Alternatively, use of serum-supplemented mammalian cell culture media as the extracting medium (Serves) to simulate more closely physiological conditions. Prepare the extracts by heating for 24 hours in an incubator preferably containing 5 ± 1% of carbon dioxide. Maintain the extraction temperature at 37±1º, because higher temperatures may cause denaturation of serum proteins.

Procedure Prepare the monolayers in 35-mm diameter plates using 2 mL of the above cell culture preparation. Aspirate the culture medium from the monolayers, and replace it with either extracts of the sample, USP Negative Control Plastic RS (to provide a Negative Control), or USP Positive Bioreaction Extract RS (to provide a Positive Control). The serum-supplemented and serum-free cell culture media extracts are tested in duplicate without dilution (100%). The Sodium Chloride Injection extract

Cytotoxicity Testing

73

is diluted with serum-supplemented cell culture medium and tested in duplicate at 25% extract concentration. Incubate all cultures for 48 hours at 37 ± 1º in an incubator preferably containing 5 ± 1% of carbon dioxide. Examine each culture at 48 hours, under a microscope, using cytochemical stains, if desired.

Interpretation of Results Proceed as directed for Interpretation of Results under Agar Diffusion Test but using Table 6.3. Repeat the test if the suitability of the system is not confirmed. The Sample meets the requirements of the test if the cultures treated with the Samples show not greater than a mild reactivity (Grade 2). If the cultures treated with the Sample show a significantly greater reaction than the cultures treated with the Negative Control, repeat the test with several quantitative dilutions of the extracts. For each of these three procedures, it should be kept in mind that while USP requires that tests be performed in duplicate, ISO requires that they be done in triplicate.

CORRELATION WITH IN VIVO RESULTS Cytotoxicity testing for medical devices is a very useful screening tool, but it must be kept firmly in mind that the correlation of results from these assays with intact animal tests (and with observed effects in humans) is poor at best. This issue was researched more than twenty years ago (Wilsnack et al., 1973; Wilsnack, 1976) with side-by-side comparisons of cytotoxicity results with those of animal tests conducted on the same samples. The results demonstrated limited correlation and “the ellipsoid effect,” the best correlation being between extreme results in the different tests (Gad, 2000). The author has also tried to correlate results of cytotoxicity and concurrent animal tests (particularly subcutaneous injection and implantation tests, where one would expect the best case) only to find that there were high levels of false negatives and positives, though predominantly the latter. So investigators are cautioned to not place too much faith and weight on the results of these assays.

CONCLUSION Cytotoxicity tests, as an initial screen for toxicities of both plastics and elastomers and of leachates from them, have been in use since the 1960s (Rosenbluth et al., 1965), but have also been recognized for a long time to be limited to effectively serving only as screens that in effect say “look at this and evaluate/consider further.” This limitation to use as screens as opposed to as definitive tests is due to the, at best, moderate correlation of their results with in vivo findings (Wilsnack, 1976). Regulatory agencies recognize the limitations of these test systems, and users should bear in mind the categorical (as opposed to truly quantitative) nature of results.

REFERENCES AAMI (1997). Biological Evaluation of Medical Devices. Association for the Advancement of Medical Instrumentation, Arlington, VA, pp. 69–82.

74

Safety Evaluation of Medical Devices

Bagley, D.M., Rizvi, P.Y., Kong, B.M. and DeSalva, S.J. (1992). Factors affecting the use of the hens’ egg chorioallantoic membrane assay as a model for eye irritation potential. Toxicol. Cut. Ocul. Toxicol., 10: 95–104. Barile, F.A. (1994). Introduction to in vitro Cytotoxicity, Mechanisms and Methods. CRC Press, Boca Raton, FL. Blein, O., Adolphe, M., Lakhdar, B. et al. (1991). Correlation and validation of alternative methods to the Draize eye irritation test (OPAL project). Toxicol. in vitro, 5: 555–557. Borenfreund, E. and Puerner, J. (1985). Toxicity determined in vitro by morphological alterations and neutral red absorption. Toxicology Letters, 24: 119–124. Borenfreund, E. and Puerner, J.A. (1987). Short-term quantitative in vitro cytotoxic assay involving an S-9 activating system. Cancer Lett., 34: 243–248. Borenfreund, E.L. and Puerner, J. (1984). A simple quantitative procedure using monolayer cultures for cytotoxicity assays (HTD/NR-90). J. Tissue Culture Methods, 9(1): 7–9. Bruner, L.H., Kain, D., Roberts, D.A. and Parker, R.D. (1991). Evaluation of seven in vitro alternatives for ocular safety testing. Fund. Appl. Toxicol., 17: 136–149. Bulich, A.A., Tung, K.K., and Scheibner, G. (1990). The luminescent bacteria toxicity test: its potential as an in vitro alternative. J. Biolumin. Chemilumen., 5: 71–77. Caldwell, J. (1993). ‘Biochemical Basis of Toxicity,’ In: B. Ballantyne, T. Marrs and P. Turner (Eds.), General and Applied Toxicology. Stockton Press, New York, pp. 169–183. CDRH (1992). Regulatory Requirements for Medical Devices. Food and Drug Administration, Rockville, MD. Gad, S.C. (2000). In vitro Toxicology, 2nd Edn. Taylor and Francis, Philadelphia, PA. ISO/DIS 10993-5 Biological Testing of Medical and Dental Materials and Devices—Part 5: Tests for Cytotoxicity: In Vitro Methods. Itagaki, H., Hagino, C., Kato, S., Kobayashi, T. and Umeda, M. (1991). An in vitro alternative to the Draize eye irritation test: Evaluation of the crystal violate staining method. Toxicol. in vitro, 5: 139–143. Kemp, R.B., Meredith, R.W.J. and Gamble, S.H. (1985). Toxicity of commercial products on cells in culture: a possible screen for the Draize eye irritation test. Food Chem. Toxicol., 23: 267–270.

North-Root, H., Yackvien, F. Demetrulias, J., Gacula, M., Jr. and Heinze, J.E. (1982). Evaluation of an in vitro cell toxicity test using rabbit corneal cells to predict the eye irritation potential of a surfactant. Toxicol. Lett., 14: 207–212. Northup, Sharon J. (1986). Mammalian cell culture models. In: Von Recum, A.F. (Ed.), Handbook of Biomaterials Evaluation, pp. 209–225. Macmillan Publishing Co., New York. Northup, Sharon J. (1987). Cytotoxicity tests of plastics and elastomers. Pharmacopoeial Forum, 13: 2939–2942. Northup, S. (1992) Cytotoxicity and mutagenicity. In: Didisheim, P. Harker, L., and Ratner, B. (Eds), Guidelines for Blood Material Interactions. Riddell, R.J., Clothier, R.H. and Balls, M. (1986). An evaluation of three in vitro cytotoxicity assays. Food Chem. Toxicol., 24: 469–477. Rohde, B.H. (1992). In vitro methods in ophthalmic toxicology. In: Chiou, G.C.Y. (Ed.), Ophthalmic Toxicology. Raven Press Ltd., New York, pp. 109–165. Rosenbluth, S.A., Weddington, G.R., Guess, W.L. and Autian, J. (1965). Tissue culture method for screening toxicity of plastic materials to be used in medical practice. J. Pharm Sci., 54: 156–159. Scaife, M.C. (1985). An in vitro cytotoxicity test to predict the oculara irritancy potential of detergent products. Food Chem. Toxicol., 23: 253–258. Shopsis, C. and Sathe, J. (1984). Uridine uptake inhibition as a cytotoxic test: correlation with the Draize test. Toxicology, 29: 195–296. Sina, J.F., Ward, G.J., Laszeh, M.A. and Gautheron, P.D. (1992). Assessment of cytotoxic assays as predicators of ocular irritation of pharmaceuticals. Fund. Appl. Toxicol., 18: 515–521. USP (1996). The United States Pharmacopoeia XXIII. The United States Pharmacopeial Convention, Rockville, MD. Wilsnack, R.E. (1976). Quantitative cell culture biocompatibility testing of medical devices and correlation to animal tests. Biomat. Med. Dev. Art. Org., 4: 235–261. Wilsnack, R.E., Meyer, F.J. and Smith, J.G. (1973). Human cell culture toxicity testing of medical devices and correlation to animal tests. Biomat. Med. Dev. Art. Org., 1: 543–562.

7 Hemocompatibility

Hemocompatibility—a lack of significant adverse interactions of a device with the formed elements of the blood—can be one of the most complex of the standard safety concerns for devices to be evaluated. Properly done for a long term cardiovascular implant device (such as a stent) as an independent entity, it could also be the most expensive of the standard, short term responses endpoint to evaluate. ISO 10993 Part 4 (Selection of Tests for Interactions with Blood) presents 25 different categories of assays for such evaluations, and FDA expectations can be yet more extensive. While in vitro methods have been used for screening materials (Motlagh et al., 2006), they have severe limitations. Few materials have consistently shown good hemocompatibility in both arterial and venous blood flow environments. Because of its complexity (Beutler et al., 1995; Colman et al., 2006) results obtained from laboratory animals may not apply to man, and results from one test system may not necessarily be correlated to those obtained from a different test system. In vitro results may not predict well what happens in vivo (Didisheim et al., 1984; Lindon et al., 1978). Any hemocompatibility statement must be linked to the intended use and conditions for which the statement is valid. Blood–material interaction can range from transient hemolysis and minimal protein adsorption to activation of coagulation, complement, and significant destruction of cells. Complicated mechanisms exist in the cardiovascular system which may interact with medical devices. Devices vary enormously in type, function, and duration of blood contact (Cooper et al., 1987; Dewangee, 1987), particularly now that combination and nanotechnology devices are increasingly moving to market. Therefore, a multidisciplinary approach to hemocompatibility testing is important. This includes in vitro static and dynamic tests, acute extracorporeal tests, tests of cardiovascular devices in appropriate animal models, and clinical studies. For most devices only the in vitro static tests are performed. Complex interactions are operative between the surfaces of devices/materials and the blood, based on both chemical and physical parameters (Zaslavsky et al., 1975), such as the rate of release of chemical moieties from a device and the nature of the blood contacting device surface. A thorough review of the normal function and structure of the hematopoietic system is a discipline in itself, and far beyond the range of this chapter. Irons (1985), Brown (1984), and Williams et al. (1995) should be consulted by those interested in the background. Figures 7.1 and 7.2 provide a very rudimentary overview of the pathways involved in the generation and differentiation of the formed elements of the blood. The FDA and ISO requirements for Hematocompatability evaluation for most devices that were not implants in the vascular system (but rather have limited duration on contact) are frequently met in the part by performing a simple in vitro hemolysis test. The ASTM guidelines (ASTM, 1987) and NIH (1985) called for a more stringent approach.

Under ISO-10993-4, externally communicating devices, with indirect or circulating contact with the blood stream, or implant devices in the vascular system must be evaluated. It is recommended that this evaluation looks at five different endpoints (test categories: thrombosis, coagulation, platelets, hematology and immunology). Table 7.1 summarizes representation tests available to evaluate each of these endpoints. Each of these categories, of course, is a potential type of interaction between the blood and materials used in devices. Devices contacting and therefore potentially having an interaction with the blood are categorized by ISO as follows.

NON-CONTACT DEVICES An example is in vitro diagnostic devices, which have no biocompatibility testing requirements.

EXTERNAL COMMUNICATING DEVICES These are devices that contact the circulating blood and serve as a conduit into the vascular system. Examples include but are not limited to those below. External communicating devices that serve as an indirect blood path include but are not limited to: cannulae, extension sets, devices for the collection of blood, devices for the storage and administration of blood and blood products (e.g., tubing, needles and bags). Indirect blood path devices are assigned the simplest testing strategy by ISO. A profile of six, relatively inexpensive, in vitro tests is recommended, one test for thrombosis, one for coagulation, one for platelet count, two hematology tests, and a complement activation panel for immunology. Optional tests may also be required by the regulatory authority and this point should be clarified with them before a testing program is initiated. External communicating devices in contact with circulating blood include but are not limited to: cardiopulmonary bypass, extracorporeal membrane oxygenators, hemodialysis equipment, donor and therapeutic apheresis equipment, devices for absorption of specific substances from blood, interventional cardiology and vascular devices, percutaneous circulatory support systems, temporary pacemaker electrodes. Circulating blood devices are assigned a somewhat more complex testing strategy, reflecting the fact that 75

76

Safety Evaluation of Medical Devices

Pluripotent stem cell pool

Lymphocyte pool

Erythropoietin Megakarocytes

Myeloblasts Proerythroblasts

Platelets

Release from marrow

Reticulocyte

Leukocytes

Erythrocytes

Figure 7.1

Diagrammatic representation of the genesis of formed blood elements arising from the bone marrow.

circulating blood must blow through the device; hence device patency becomes an issue. Again, tests from the five basic categories are recommended by ISO. Additional, optional tests are also listed.

inferior vena cava filters, stents, arteriovenous shunts, blood monitors, internal drug delivery catheters, pacemaker electrodes, intravascular membrane oxygenators (artificial lungs).

IMPLANT DEVICES These are devices that are placed largely or entirely within the vascular system. Examples include but are not limited to: mechanical or tissue heart valves, prosthetic or tissue vascular grafts, circulatory support devices (ventricular-assist devices, artificial hearts, intra-aortic balloon pumps),

The ISO guidelines also provide detailed guidance as to tests to be performed for each of these types of devices. This guidance is summarized in Tables 7.3–7.7. Table 7.2 provides a codex for the significant aberrations utilized in Tables 7.2–7.7. Not covered in these tables are the specialized cases associated with cardiovascular devices.

Leukocytes

Phagocytes

Immunocytes

Lymphocytes Granulocytes

Monocytes

Neutrophils Basophils Eosinophils Macrophages (liver, spleen, bone marrow)

Figure 7.2

T Cells (thymus cell-mediated immunity)

Diagrammatic representation of the differentiation of leukocytes into white blood cells.

B Cells (bone marrow, secrete antibodies IgG, IgA, etc.)

Chapter 7:

Table 7.1

Hemocompatibility

77

ISO 10993-4: Selection of Tests for Interaction with Blood

Representative Tests Methods Test Categories

Level I

Level II

Comments

Thrombosis

Light microscopy (adhered platelets, leukocytes, aggregates, erythrocytes, fibrin, etc.) Partial thromboplastin time (nonactivated) Platelet count Leukocyte count and differential; hemolysis (plasma hemoglobin)

SEM

Light microscopy can be replaced by scanning electron microscopy if the nature of the material presents technical problems for light microscopy.

C3a, C5a, RCC Bb, iC3b, C4d, SC5b-9

Cytokines & mRNA’s

Coagulation Platelets Hematology

Immunology

FPA, D-Dimer, etc. PF-4, thromboxane B2 Reticulocyte count

STANDARD TESTS Among the wide range of tests described in Tables 7.3–7.7, there are a number which are most commonly performed. Theses are available at most contract research organizations and hospitals, and include (besides the simple tests presented here) determination of the numbers and types of formed elements of the blood (Lewis et al., 1990) and other end points such as the absorption of specific formed elements of the blood (such as lymphocytes) on the surfaces

Table 7.2

Abbreviations

Abbreviation

Meaning

Bb β-TG C-4d C-3a, C5a D-Dimer

Product of alternate pathway complement activation Beta-thromboglobulin Product of classical pathway complement activation (Active) complement split products from C3 and C5 Specific fibrin degradation products (F XIII cross-linked fibrin) Extracorporeal membrane oxygenator Electron microscopy Fibrin/fibrinogen degradation products Fibrinopeptide A Prothrombin activation fragment 1 + 2 Product of central C complement activation Interleukin-1 Inferior vena cava Magnetic resonance imaging Monoclonal antibody which recognizes the activated form of platelet surface glycoprotein IIb/IIIa Positron emission topography Platelet factor 4 Prothrombin time Partial thromboplastin time Radioimmunoassay Monoclonal antibody which recognizes the alpha granule membrane component GMP140 exposed during the platelet release reaction Product of terminal pathway complement activation Thrombin-antithrombin complex Terminal complement complex Thrombin time von Willebrand factor

ECMO EM FDP FPA F1+2 iC3b IL-1 IVC MRI PAC-1 PET PF-4 PT PTT RIA S-12

SC5b-9 TAT TCC TT VWF

Hemolysis is regarded as an especially significant screening test to perform in this category because of its measurement of red blood cell membrane fragility in contact with materials and devices. The method used should be one of the normative standard test methods for hemolysis. A panel including the last four tests encompasses the various complement activation pathways

of polymers and ceramics (Yokoyama et al., 1986). These may require more specialized equipment and approaches than are generally available.

Hemolysis Tests The simplest and most commonly conducted menocompatibility test is the in vitro homolysis test. In the direct contact hemolysis test the intact test article is placed in a solution of saline and a small amount of whole red blood is added. After a period of incubation (one hour) at 37ºC the supernate is decanted and assayed for hemoglobin. The concentration of hemoglobin is proportional to the number of red cells that were lysed. The percent hemolysis is calculated by the equation below: Percent hemolysis =

A−B × 100 C

where A is the absorbance of the test sample, B is the absorbance of a negative control, and C is the absorbance of a positive control. Hemolysis tests evaluate the acute in vitro hemolytic properties of materials, especially those intended for use in contact with blood. The concentration of substances which produces hemolysis is generally higher than that needed to produce a cytotoxic effect. The result of hemolysis testing can be correlated with acute in vivo toxicity tests. A hemolysis test is rapid, requires simple equipment, gives easily interpretable quantitative results, and can be performed in the presence of the material or on the extract. The results are compared to the controls and expressed as percent hemolysis. The average life-span of human red cells is 120 days. When the life-span is shortened, whatever the cause, there is said to be a hemolytic process, and when the marrow fails to replace the lost cells quickly enough then a hemolytic anemia reduction in red blood cell count develops. The term “hemolytic” is rather misleading, as it implies actual lysis or bursting of the red cells in the circulation. Sometimes this does occur, and then it is known as intravascular hemolysis. More often, however, the cells are damaged or in some way inadequate, and are then removed

78

Safety Evaluation of Medical Devices

Table 7.3

External Communicating Devices—Level 1—Blood Path, Indirect

Test category

Method

Comments

Thrombosis

Light microscopy (adhered platelets, leukocytes, aggregates, erythrocytes, fibrin, etc. PTT (non-activated) Platelet count Leukocyte count and differential: hemolysis (plasma hemoglobin)

Light microscopy can be replaced by scanning EM if the nature of the material presents technical problems for light microscopy

Coagulation Platelets Hematology

Immunology

C3a, C5a, TCC, Bb, iC3b, C4d, SC5b-9

Hemolysis is regarded as an especially significant screening test to perform in this category because of its measurement of red blood cell membrane fragility in contact with materials and devices. The method used should be one of the normative standard test methods for hemolysis. A panel including the last four tests encompasses the various complement activation pathways.

Level 2 (Optional) Thrombosis Coagulation Platelets

Table 7.4

Scanning EM Coagulation factor assays, including: FPA, D-dimer, F1+2, PAC-1, s-12, TAT PF-4, β-TG, thromboxane B2, 111In-labelled platelet survival

External Communicating Devices—Level 1—Circulating Blood

Test Category

Method

Comments

Thrombosis

Per cent occulusion; flow reduction; gravimetric analysis (thrombus mass); light microscopy (adhered platelets, leukocytes, aggregates, erythrocytes, fibrin, etc.); pressure drop across device PTT (non-activated) Platelet count; platelet aggregation; template bleeding time Leukocyte count and differential: hemolysis (plasma hemoglobin)

Light microscopy can be replaced by scanning EM if the nature of the material presents technical problems for light microscopy Pressure drop not recommended for devices intended for PR

Coagulation Platelets Hematology

Immunology

Table 7.5

C3a, C5a, TCC, Bb, iC3b, C4d, SC5b-9

External Communicating Devices—Level 2—Optional

Test Category

Method

Thrombosis

Scanning EM (platelet adhesion and aggregation; platelet and leukocyte morphology; fibrin) Specific coagulation factor assays; FPA, D-dimer, F1+2, PAC-1, S-12, TAT PF-4, β-TG; thomboxane B2; gamma imaging of radiolabeled platelets 111 In-labeled platelet survival Reticulocyte count; activation specific release products of peripheral blood cells (i.e., granulocytes) C3a, C5a, TCC, Bb, iC3b, C4d, SC5b-9

Coagulation Platelets

Hematology Immunology

Hemolysis is regarded as an especially significant screening test to perform in this category because of its measurement of red blood cell membrane fragility in contact with materials and devices. The method used should be one of the normative standard test methods for hemolysis A panel including the last four tests encompasses the various complement activation pathways

from the circulating blood by macrophages in the spleen in the usual way. This process is known as extra vascular hemolysis. One of the most common tests for hemolysis is the osmotic fragility test. The membranes surrounding most cells in animal tissues are semi-permeable, which means that they allow the passage of water, but prevent the passage of dissolved substances. When two solutions of different concentration are separated by such a membrane, water passes from the more dilute solution to the more concentrated one, until the

Comments

111

In-labeling is recommended for PR only

A panel including the last four tests encompasses the various complement activation pathways.

concentration on both sides is equal. This tendency for water to flow in one direction is called osmosis, and the pressure exerted as it does so is known as the osmotic pressure. Two solutions of equal concentration are known as “isotonic.” For human blood this is equivalent to 0.9% saline solution. When their concentrations are unequal the more dilute solution is “hypotonic (hypo = low) and the more concentrated solution is “hypertonic” (hyper = over, above). If red cells are to retain their shape and function properly, the hemoglobin solution inside the cell and the plasma

Chapter 7:

Table 7.6

Method

Thrombosis

Per cent occlusion; flow reduction; autopsy of device (gross and microscopic); autopsy of distal organs (gross and microscopic) PTT (non-activated), PT, TT; plasma fibrinogen, FDP Platelet count; platelet aggregation Leukocyte count and differential; hemolysis (plasma hemoglobin)

Platelets Hematology

Immunology

Comments

C3a, C5a, TCC, Bb, iC3b, C4d, SC5b-9

outside the cell must be isotonic. Similarly when dealing with red cells in the laboratory, solutions must be isotonic with the contents of the cell.

The Osmotic Fragility Test The degree of hemolysis of cells in hypotonic solutions depends largely on their shape. Cells which are already spherocytic (spherical) are easily lysed, that is, they are more fragile. Flattened cells on the other hand, are more resistant to lysis than normal cells. A series of solutions are prepared containing salt concentrations from 0.3 to 0.6 g per 100 mL. A volume of blood is then added to each of these, and after half an hour the degree of hemolysis is found by spinning down the intact red cells and measuring the color intensity of the supernatant. A normal control blood must always be treated in the same way for the purpose of comparison, and to check the quality of the reagents (Kirk et al., 1975). The process proceeds as follows: 1. Method: (a) Place two rows of seven 10 mL tubes in a rack, and label one row T1–T7 (test) and the other row C1–C7 (control). (b) Label another four tubes T. Std., T. Blank, C. Std., and C. Blank. Place these in the appropriate places in the rack. (c) Using one row only, set up tubes as in the table below. Table 7.7

Implant Devices—Level 2—Optional

Test Category

Method

Thrombosis Coagulation

Scanning EM; angiography Specific coagulation factor assays; FPA, D-dimer, F1+2, PAC-1, S-12, TAT 111 In-labeled platelet survival PF-4, β-TG; thomboxane B2; gamma imaging of radiolabeled platelets Reticulocyte count; activation specific release products of peripheral blood cells (i.e., granulocytes) IL-1 and other cytokines; detection of messenger-RNA specific for cytokines

Platelets

Hematology

Immunology

79

Implant Devices — Level 1

Test Category

Coagulation

Hemocompatibility

Comments

Hemolysis is regarded as an especially significant screening test to perform in this category because of its measurement of red blood cell membrane fragility in contact with materials and devices. The method used should be one of the normative standard test methods for hemolysis. A panel including the last four tests encompasses the various complement activation pathways.

Tube

1

2

3

4

5

6

7

1% NaCl (mL) Distilled water (mL) Final NaCl concn. (g%)

3.0 7

3.5 6.5

4 4.5 6.0 5.5

0.3

0.35 0.4 0.45 0.5

Std. Blank

5.0 5.0

5.5 4.5

6.0 — 4.0 10

10

0.55 0.6

(d) Mix the solutions carefully, then transfer 5 mL from each tube to the corresponding tube of the second row. This ensures that each concentration is the same for test and control. (e) To the test row add 0.05 mL of control blood in the same way. (g) Mix all the tubes and allow them to stand at room temperature for 30 min. (h) Mix them again, then centrifuge all the tubes at 3000 rev/min for 5 min. (i) Using the appropriate blank, read the color intensities of the test row in a colorimeter, using an Ilford 625 green filter. Take care not to disturb the red cell layer when transferring the clear supernatants to the cuvettes for reading. (j) Repeat the procedure with the control row, taking care to change the blank. 2. Calculation and results: One now has two sets of eight readings, including the standards. The standards, which are cells in water, represent 100% hemolysis. They are different from each other, because the hemoglobin levels of the two bloods are different. Using one set of readings, calculate the percentage hemolysis in each tube. Percentage hemolysis =

Test reading × 100 Standard reading

Calculate the second set of results in the same way. The results are best expressed in the form of a graph showing percentage hemolysis against the NaCl concentrations. 3. Factors influencing the results: (a) The blood must be as fresh as possible, preferably less than 4 hours old. Cells deteriorate on standing and begin to lyse spontaneously. (b) Defibrinated or heparinized blood is most suitable for this test, as such blood would not contain any

80

Safety Evaluation of Medical Devices

extra salts. Blood anticoagulated with EDTA is used also. (c) The ratio of blood to saline affects the results; a ratio of 1: 100 is usually used. (d) The pH of the saline also affects the results. In order to standardize this, a stock solution of buffered saline may be prepared which is osmotically equivalent to 10% NaCl: Sodium chloride (NaCl) 178 g Disodium hydrogen phosphate 27.21 g (Na2PO4.2H2O) Sodium dihydrogen phosphate 4.86 g (NaH2PO4.2H2) Distilled water to 2 liters This solution is diluted 1 in 10 just before use. 4. Normal range Slight hemolysis: 0.45–0.4 g% NaCl Complete hemolysis 0.35–0.3 g% NaCl

Erythrocyte Stability The erythrocyte stability test provides a sensitive measure of the interaction of extractable or leachable substances with the plasma membrane of erythrocytes and is reflected as changes in the osmotic fragility of the erythrocytes. This test can detect leachables at concentrations slightly below the sensitive levels of many cytotoxicity systems. Hypotonic saline or distilled water (as described earlier under hemolysis) can be adjusted to the required tonicity. Extractors are adjusted to give osmolarity appropriate to hemolyze about 50% of the erythrocytes. Usually rabbit blood is used (though human blood is preferred), diluted with isotonic saline to about 1% hematocrit. One-tenth of a mL of this stock erythrocyte solution is added to 5 mL of hypotonic extract, and the surviving cells are counted. The relative hemolysis, number of cells lysed in the extract versus number of cells lysed in the control, is reported. By performing the tests with a series of dilutions of the extract, the concentration of extract at which no detectable change occurs can be established and compared with data from other materials or from extracts prepared under different conditions. Cell size distribution profiles can also be obtained in this test, giving an indication of the degree of swelling or morphologic changes.

Whole Blood Clotting Time Whole blood clotting time may be measured by modifying the Lee-White method or other relevant tests. The Leo White test measures the recalcification as an indicator of coagulability (Kretschner et al. 2004). Such measures, however, are influenced by blood or sample dilution. Thrombin time, prothrombin time, and/or platelet counts should be included. • partial thromboplastin time (PTI) — shortened time = activation of intrinsic pathway — sensitive to all know clotting factors except VII and XII • prothrombin time (PT) — measures activation of extrinsic pathway — measures time required for recalcified plasma to clot in presence of thromboplastin • thrombin time (TT) — useful in detecting inhibitors of the thrombin– fibrinogen reaction — measure the availability of functional fibrinogen.

Thrombogenicity Testing for thrombogenicity is normally done by examining platelet and fibrinogen turnover and observing thrombus formation and resulting emboli. Because thrombogenicity tests are usually difficult, controversial, and expensive, manufacturers should consult with the FDA to verify the proper model and test protocol (as briefly described below). Thrombosis is the formation or existence of a blood clot within the vascular system. When associated with a device, it can be a life-threatening event because the clot, called a thrombus, can occlude a vessel and stop the blood supply to an organ or body part. If detached, the thrombus becomes an embolus and may occlude a vessel at a distance from the original site. When measuring thrombosis, the test endpoint is the size of clot formation or the adherence of platelets, leukocytes, erythrocytes, or other aggregates on the test device. In the light microscopy method, an intact sample may be exposed to whole blood ex vivo. Ex vivo means away from or outside of the body and, in ex vivo experiments, some of an animal’s blood is caused to bypass the normal circulatory system and pass through or across a device, then flow back into the animal’s body. Light microscopy is used to scan the material for evidence of thrombus formation. Alternatively, the material or device may be excised after a suitable period of exposure in vivo, then scanned for thrombus formation using a light microscope. Thrombosis “tests” give a yes or no answer. There is either thrombus formation or there is not. Percent occlusion, flow reduction and gravimetric analysis are all attempts to quantify the amount of thrombus formation. Percent occlusion is visually assessed after a device has been in use and has been removed. Percent occlusion is a measure of the severity of the thrombotic process in a conduit. Flow reduction is a measure of the drop in rate or volume of blood flow through a device after a period of implantation. Gravimetric analysis is a weight measure of thrombus mass after removal of the mass from a device after a period of use. Scanning electron microscopy is a method of visually assessing a device on a micron scale. When used on explanted materials or devices it may give a closer visual assessment of thrombus formation, capsular formation, or device performance. Angiography (an X-ray of blood vessels which have been made radio-opaque by the injection of a dye) is a method of taking an X-ray of the vasculature following injection of a radiopague substance to obtain a description of the blood vessels or the arterial pulse.

Complement Activation Inappropriate use of excess complement activation may lead to unwanted tissue damages or cause cardiopulmonary distress in patients (Henderson, 1989). Complement activation is usually measured by the conversion of C3 to C3a and/or C5 to C5a. The hemolytic complement expressed in CH50 is generally not sensitive enough to detect complement activation caused by biomaterials and is not acceptable as a complement activation assay. The classical complement system consists of nine separate protein components (numbered C1 through C9) acting in sequence. When activated, complement components interact sequentially with one another in a cascade. Activation of some complement components results in the cleavage of a component into two fragments. In some cases, the larger fragments join other activated fragments and the smaller fragments, such as C3a and C5a, have inflammatory properties.

Chapter 7:

The C3a and C5a cause vasodilation and increase capillary permeability.

Protein Adsorption The adsorption of plasma protein is generally the first event that occurs when blood contacts a foreign surface (Lemm and Unger, 1980). This protein layer has a great influence on the thrombogenicity of a material. One of the more commonly used techniques is the radiolabeling of protein with 125I. The measurements consist of three steps: • the exposure of a solid surface to a solution containing the radiolabeled proteins, • rinsing to remove all but the adsorbed protein, and • measurement of the radioactivity retained by the surface. This technique provides a direct measurement of the amount of protein adsorbed on a surface. Protein adsorption can also be studied from flowing solutions in specially designed flow chambers. Recently, real-time spectrophotometric measurements of dynamic protein adsorption have been done by Fourier and transformed into infrared-attenuated total reflectance.

Coagulation Coagulation refers to the process of blood clotting, which results from the initiation of a cascading enzymatic pathway where the product on one reaction is an enzyme which catalyzes another, subsequent, reaction. The outcome of coagulation is the formation of a clot or thrombus. When measuring coagulation, the test endpoint is enzyme activation or suppression (not thrombus formation). Clotting time, as designed by Lee and White, is probably the earliest coagulation test developed. It is not discussed in ISO 10993 Part 4, although it is still frequently used to screen materials for blood compatibility. A sample of blood is removed from an animal exposed to material or device. The time at which the blood is withdrawn is noted as accurately as possible. The syringe is emptied into a small glass tube, which is rotated endwise every 30 seconds. The point at which the blood no longer flows from its position but maintains its surface contour when inverted is taken as the endpoint. Normal clotting time in humans is about 6.5 minutes. Thromboplastin is the third blood coagulation factor (Factor III). Partial thromboplastin time (PTT) is the clotting time of recalcified citrated plasma upon the addition of partial thromboplastin obtained from mammalian brain or lung. Shortening of the PTT following contact with a material indicates activation of coagulation factors; a prolonged PTT suggests a deficiency. A blood sample, as citrated plasma, is obtained from an animal that has been exposed to the intact material. An excess of calcium ions and thromboplastin are added and the time to clotting measured. Prothrombin is a circulating protein which, when acted upon by thrombokinase, forms thrombin. Prothrombin time (PT) is related to prothrombin concentration and the accessory factors, Factor V, Factor VII, and Factor X. In the presence of thromboplastin, clotting time depends on the concentrations of these four factors. A blood sample is obtained from an animal that has been exposed to the intact material. An excess of calcium ions and thromboplastin is added and the time to clotting measured. A prolonged prothrombin time indicated a deficiency of prothrombin, Factors V, VII, X, or fibrinogen, indicating the implant has inactivated, absorbed or otherwise interfered with the concentration of these proteins.

Hemocompatibility

81

Thrombin is a protein found in shed blood. Formed from prothrombin, it reacts with soluble fibrinogen, converting it to fibrin which forms the basis of blood clots. Thrombin time (TT) is the time required for plasma to clot when a solution of thrombin is added. A blood sample as plasma is obtained from an animal that has been exposed to the intact material. A solution of thrombin is added to the plasma and the time to clotting measured. A prolonged thrombin time indicates a deficiency in fibrinogen. Plasma fibrinogen is a protein in the blood which, when acted upon by thrombin and calcium, forms fibrin. A sample of plasma is obtained from an animal that has been exposed to the intact material. Active fibrinogen is measured indirectly by using a commercially available thrombin time assay. Thrombin time is dependent on fibrinogen and can be an accurate measure of its active concentration. Fibrin/fibrinogen degradation products are byproducts of degraded fibrin and/or fibrinogen. A sample of plasma is obtained from an animal that has been exposed to the intact material. An immunoassay is performed by exposing the plasma to fibrin/fibrinogen antibodies per the instructions in commercially available tests. Specific coagulation factor assays for fibrinopeptide A, D dimer (a fibrin degradation product), F1+2 (prothrombin activation fragment 1+2), PAC-1 (monoclonal antibody which recognizes the activated form of platelet surface glycoprotein IIb/IIIa), S-12 (monoclonal antibody which recognizes the alpha granule membrane component 9GMP140 exposed during the platelet release reaction), or TAT (thrombin-antithrombin complex) may be performed on blood samples taken from animals exposed to intact, implanted material.

PLATELETS Platelet count—Platelets are flat, round cells found in the circulating blood. They play an important role in blood coagulation, hemostasis, and thrombus formation. When a small vessel is injured, platelets adhere to each other and the edges of the injury and form a plug. The plug or blood clot soon retracts and stops the loss of blood. A blood sample is obtained from an animal that has been exposed to the intact material and the number of platelets per mm3 determined. Normal human values are 200,000 to 300,000. Platelet aggregation is induced when cells at the site of injury secrete epinephrine, or when collagen, thrombin or other agents are produced at the site. Platelet aggregation can by induced in vitro by the addition of these agents exogenuously. To evaluate the ability of platelets to aggregate, plasma is placed in a beaker and the exogenous agents added with constant stirring. As the platelets aggregate the plasma becomes progressively clearer. An optical system (aggregometer) is used to detect the change in light transmission. Delayed or reduced platelet aggregation, or spontaneous aggregation, is a sign of platelet activation. Assays for PF-4 (platelet factor 4), β-TG (beta-thromboglobulin), or thromboxane B2 may be performed on blood samples taken from animals exposed to intact, implanted material. Gamma imaging of radiolabeled platelets performed on 111Indium-labelled platelet survival times may be determined in situ in animals exposed to intact, implanted material.

82

Safety Evaluation of Medical Devices

CONCLUSION Though hematocompatability has long been identified as a concern for medical devices and biomaterials (Mason, 1972; Autian, 1977; Wilsnack and Bernadyn, 1979), it is only recently that the standards for evaluation of the relevant end points to utilize available technology (ISO-10993-4). Currently, for materials and for devices with limited (in either extent of duration) exposure to the circulated blood, a limited battery of in vitro evaluations, as described in Table 7.1, should be adequate to ensure hematocompatability. For devices with extended contact with circulating blood (systemic circulation), however, a much more extensive evaluation, including at least an in vivo study in a suitable model species should be considered.

REFERENCES ASTM F 756-87 (1987). Standard Practice for Assessment of Hemolytic Properties of Materials. Autian, J. (1977). Toxicological evaluation of biomaterials: primary acute toxicity screening program. Artif. Org., 1, 53–60. Brown, B.A. (1984). Hematology: Principles and Procedures. Lea and Febiger, Philadelphia. Beutler E., Lichtman, M.A., Coller, B.S. and Kipps, T.J. (1995). Williams Hematology, 5th edn. McGraw-Hill, New York. Colman, R.W., Marder, V.J., Clowes, A.W., George, J.N. and Goldhaber, S.Z. (2006). Hemostasis and Thrombosis: Basic Principles and Clinical Pratice, 5th edn. Lippincott Williams and Wilkins, Philadelphia, PA. Cooper, S.L., Fabrizius, D.J. and Grasel, T.G. (1987). Methods of assessment of thrombosis ex vivo. In: Leonard, E.F., Turitto, V.T., and Vroman, L. (Eds.), Blood in Contact with Natural and Artificial Surfaces. Ann. N.Y. Acad Sciences, 516: 572–585. Dewangee, M.K. (1987). Methods of assessment of thrombosis in vivo, In: Leonard, E.F., Turitto, V.T. and Vroman, L. (Eds.), Blood in Contact with Natural and Artificial Surfaces. Ann N.Y. Acad Sciences, 516: 541–571. Didisheim, P., Dewanjee, M.K., Kaye, M.P., Frisk, C.S., Fass, D.N., Tirrell, M.V. and Xillman, P.E. (1984). Nonpredictability of long-term

in vivo response from short-term in vitro or ex vivo blood/material interactions. Trans. Am. Soc. Artif. Int. Organs, 30: 370–376. Henderson, L.W. (1989). Immunotoxicology of blood–synthetic membrane Interactions. Fund. Appl. Toxicol., 13: 228–234. Irons, R.D. (Ed). (1985). Toxicology of the Blood and Bone Marrow. Raven Press, New York. ISO (2002). ISO-10993-Part 4 Selection of Tests for Interactions with Blood. Kirk, C.J.C., Peel, R.N., James, K.R., and Kershaw, Y. (1975). Basic Medical Laboratory Technology. John Wiley and Sons, New York. Kretschmer, V., Daraktchiev, A., Bade, S., Karger, R. and Kratzer, M.A.A.(2004). Does hemodilution enhance coagulability?, AINS, Anästhesiol. Intensivmed. Notf. Med. Schmerzther, 39: 751–756. Lemm, N. and Unger, V. (1980). Adsorption of blood proteins on different polymer surfaces in vitro. In: Winter, G.D., Leary, J.L. and deGroot, K. (Eds.), Evaluation of Biomaterials. John Wiley, New York. Lewis, S.M., Rowan, R.M., and Kubota, F. (1990). Recommended methods for the visual determination of white cell and platelet count, J. Clin. Pathol, 43: 932–936. Lindon, N.N., Rodvien, R., Brier, D. et al. (1978). In vitro assessment of interaction of blood with model surfaces. J. Lab. Clin. Med., 92: 904–914. Mason, R.G. (1972). Some methods of in vitro estimation of the blood compatibility of biomaterials. Bull. N.Y. Acad. Med., 48(2): 407–424. Motlagh, D, Yang, J., Lui, K.Y., Webb, A.R. and Ameer, G.A. (2006). Hemocompatibility evaluation of poly(glycerol-sebacate) in vitro for vascular tissue engineering. Biomaterials, 27: 4315–4325. NIH (1985). Guidelines for blood-materials interactions. Report of the National Heart, Lung, and Blood Institute-Working Group, NIH Publication No. 85-2185, September, 1985. Tripartite Biocompatibility Guidance for Medical Devices (1986). Prepared by toxicology sub-group of the Tripartite Subcommittee on medical devices, September 1986. Wilsnack, R.E. and Bernadyn, S.A. (1979). Blood compatibility of medical device materials as measured by lymphocyte function. Biomat. Med. Dev. Artif. Org., 7(4): 527–546. Yokoyama, M., Nakahashi, T., Nishimura, T., Maeda, M., Inoue, S., Kataoka, K. and Sakurai, Y. (1986). Adhesion behavior of rat lymphocytes to poly (ether)–poly (amino acid) block and graft copolymers. J. Biomed. Mater. Res., 20: 867–878. Zaslavsky, B.Y., Ossipov, N.N. and Rogozhin, S.V. (1978). Action of surface-active substances on biological membranes. Ill. Comparison of hemolytic activity of ionic and nonionic surfactants. Biochim. Biophys. Acta, 510: 15–159.

8 Local Tissue Tolerance

Local tissue tolerance or irritation studies assess the short-term and generally localized hazards of medical devices in the immediate region of their tissue contact. Topical local (tissue) tolerance effects are almost entirely limited to irritation. Though this usually means dermal irritation, it can also be vaginal, muscular, vascular, mucous membrane, rectal, nasal, or ocular. All of these but ocular irritation use some version of a common subjective rating scale (see Table 8.1) to evaluate responses. Most commonly recognized is the use of this scale, in the primary dermal irritation (PDI) test, which is performed for those agents that are to be administered to patients by application to the skin. As with all local tolerance tests, it is essential that the material be evaluated in “condition of use”— that is, in the final product, applied to test animals in the same manner that the device or biomaterial is to be used clinically. If appropriate (under applicable regulations) or necessary due to the nature or mode of use of the device, an extract can be evaluated. Such extracts are generally evaluated in the intracutaneous reactivity test. Additionally, for devices, there is a requirement to evaluate the potential pyrogenicity of the final product (and generally, on a quality control basis of subsequent appropriately selected samples of production lots) if: ● ●

●

the dermal irritation caused by a substance is investigated by observing changes ranging from erythema and edema to ulceration produced in rabbit skin when irritants are applied. These skin reactions are produced by diverse physiologic mechanisms, although they are easily observed visually and by palpitation. Evaluation of materials for their potential to cause dermal irritation and corrosion due to acute contact has been common for industrial chemicals, cosmetics, agricultural chemicals, and consumer products since at least the 1930’s (generally, pharmaceuticals are only evaluated for dermal effects if they are to be administered topically—and then by repeat exposure tests, which will not be addressed here). As with acute eye irritation tests, one of the earliest formal publications of a test method (though others were used) was that of Draize et al. in 1944 (Geller et al., 1985; SOT, 1989). The methods currently used are fundamentally still those proposed by Draize et al. and, to date, have changed very little since 1944. Efforts have been underway for some 20 years to develop alternatives that either don’t use animals, or are performed in a more humane and relevant (to human exposure) manner. Among the most fundamental assessments of the safety of a product or, indeed, of any material that has the potential to be in contact with a significant number of people in our society, are tests in animals which seek to predict potential skin irritation or corrosion. Like all the other tests in what is classically called range-finding, tier I, or acute battery, the tests used here are both among the oldest designs and are currently undergoing the greatest degree of scrutiny or change. Currently, all the established test methods for these endpoints use the same animal model, the rabbit (almost exclusively the New Zealand White), though some other animal models have been proposed. Testing is performed to evaluate the potential occurrence of two different, yet related, endpoints. The broadest application of these is an evaluation of the potential to cause skin irritation, characterized by erythema (redness) and edema (swelling). Severity of irritation is measured in terms of both the degree of these two parameters and how long they persist. There are two types of irritation tests, each designed to address a different concern.

the label claims that the device is pyrogen free, the device comes in contact with blood or spinal fluid, and/or the device is an intraocular lens.

DERMAL IRRITATION Skin irritation testing is performed to demonstrate the irritation potential of the device, i.e., for initiating or aggravating damage through its contact with the skin (Draize, 1955, 1959). Primary skin irritation is usually done according to the regulations of the Consumer Product Safety Commission, Title 16, Chapter II, Part 1500, or some variation thereof (such as ISO and ASTM, 1991). The purpose of the study is to determine the dermal irritation potential of the test article to the intact and abraded skin of the rabbit (the latter to stimulate wound tissue). Skin absorption occurs through a process of binding, partitioning, and diffusion of test materials on and into the skin. Penetration has been assessed in vivo by measuring at different times the amount of test substances at different layers of the skin. Blood levels of the test sample have been measured in this test. A complicated series of chemical and physiological responses result in primary skin irritation. When skin is exposed to toxic substances, a modified form or the Draize rabbit skin test (fewer animals are now used), first outlined by John Draize in 1944, remains an important source of safety information for government and industry. In this test,

(a) Primary (or acute) irritation, a localized reversible dermal exposure response resulting from a single application of, or exposure to, a chemical without the involvement of the immune system. (b) Cumulative irritation, a reversible dermal response which results from repeated exposure to a device (each individual exposure possessing no or limited potential to causing acute irritation). Most regulations and common practice characterize an irritation that persists 14 days past the end of exposure as other than reversible. The adult human has 1.8 m2 of skin, varying in thickness from 0.02 inches on the eyelids to 83

84

Safety Evaluation of Medical Devices

Table 8.1 Studies Method

Evaluation of Local Tissue Reactions in Tissue Irritation Basis/Endpoint

Erythema and eschar formation No erythema Very slight erythema (barely perceptible) Well-defined erythema Moderate to severe erythema Severe erythema (beet redness) to slight eschar formation (injuries in depth) Necrosis (death of tissue) Eschar (sloughing and scar formation) Edema formation No edema Very slight edema (barely perceptible) Slight edema (edges of area well-defined by definite raising) Moderate edema (raised approximately 1 mm) Severe edema (raised more than 1 mm and extending beyond the area of exposure) Total possible score for primary irritation

Reference 0 1 2 3 4 +N +E 0 1 2 3 4 8

Source: Draize et al. (1944).

0.12–0.16 inches on the back, palms, and soles of the feet (Hipp, 1978). The epidermis, the outer portion of the skin, is several layers thick, covers the entire surface of the body, and is referred to as the horny layer or stratum corneum. It is the first line of defense against physical, chemical, and thermal exposure. The skin is host to normal bacterial flora consisting of Micrococci and Corynebacterium, which play an important role in the protection against infection. The melanocyte system, responsible for skin colonization, is located at the interface of the epidermis and the dermis. New cells are constantly being formed from the basal layer and slowly migrate to the surface, replenishing themselves approximately every two weeks (Monash and Blank, 1958; Matoltry et al., 1968). Irritation is generally a localized reaction resulting from either a single or multiple exposure to a physical or chemical entity at the same site. It is characterized by the presence of erythema (redness), edema, and may or may not result in cell death. The observed signs are heat (caused by vessel dilation and the presence of large amounts of warm blood in the affected area), redness (due to capillary dilation), and pain (due to pressure on the sensory nerves). The edema often observed is largely due to plasma, which coagulates in the injured area, precipitating a fibrous network to screen off the area, thereby permitting leukocytes to destroy exogenous materials by phagocytosis. If the severity of injury is sufficient, cell death may occur, thereby negating the possibility of cellular regeneration. Necrosis is a term often used in conjunction with cell death, and is the degeneration of the dead cell into component molecules which approach equilibrium with surrounding tissue (Montagna, 1961).

Primary Dermal Irritation Test A. Rabbit Testing Procedure 1. A group of at least five New Zealand white rabbits are screened for the study. 2. All rabbits selected for the study must be in good health; any rabbit exhibiting sniffles, hair loss, loose stools, or apparent weight loss is rejected and replaced. 3. One day (at least 18 hr) prior to application of the test substance, each rabbit is prepared by clipping the hair from the back and sides using a small animal clipper.

A size No. 10 blade is used to remove long hair and then a size No. 40 blade is used to remove the remaining hair. 4. Three animals with skin sites that are free from hyperemia or abrasion (due to shaving) are selected. Skin sites that are in the telogen phase (resting stage of hair growth) are used; those skin sites that are in the anagen phase (stage of active growth, indicated by the presence of a thick undercoat of hair) or not used. B. Study Procedure 1. As many as four areas of skin, two on each side of the rabbit’s back, can be utilized for sites of administration. 2. Separate animals are not required for an untreated control group. Each animal serves as its own control. Indeed, up to eight separate sites may be used for a single rabbit. 3. The intact (free of abrasion) sites of administration are assigned a code number. Typically a suitably prepared extract from a device or biomaterial and an extraction solution (“vehicle” control) are tested on each animal. 4. Application sites should be rotated from one animal to the next to ensure that the test substance and controls are applied to each position at least once. 5. Each test or control substance is held in place with a 1′′× 1′′ 12-ply surgical gauze patch. The gauze patch is applied to the appropriate skin site and secured with 1”-wide strips of surgical tape at the four edges, leaving the center of the gauze patch non-occluded. 6. If an extraction solution is being evaluated, a patch is applied and secured to the appropriate skin site. A 1-mL tuberculin syringe is used to measure and apply 0.5 mL of test substance to the patch. 7. The negative control site is covered with an untreated 12-ply surgical gauze patch (1′′ × 1′′). 8. The entire trunk of the animal may be covered with an impervious material (such as Saran Wrap®) for a 24-h period of exposure. The Saran Wrap® is secured by wrapping several long strips of athletic adhesive tape around the trunk of the animal. The impervious material aids in maintaining the position of the patches and retards the evaporation of volatile test substances. Alternatively, a Hilltop chamber type self enclosed applicator may be utilized. 9. An Elizabethan collar is fitted and fastened around the neck of each test animal. The collar remains in place for the 24-hr exposure period. The collars are utilized to prevent removal of wrappings and patches by the animals, while allowing the animals food and water ad libitum. 10. The wrapping is removed at the end of the 24-hr exposure period. The test substance skin site is wiped to remove any test substance still remaining. When colored test substances (such as dyes) are used, it may be necessary to wash the test substance from the test site with appropriate solvent or vehicle (one that is suitable for the substance being tested). This is done to facilitate accurate evaluation for skin irritation. 11. Immediately after the removal of the patches, each 1′′ × 1′′ test or control site is outlined with an indelible marker by dotting each of the four corners. This procedure delineates the site for identification. C. Observations 1. Observations are made of the test and control skin sites 1 hr after removal of the patches (25 hr post-initiation

Chapter 8:

of application). Erythremia and edema are evaluated and scored on the basis of the designated values presented earlier in Table 8.1. 2. Observations are again performed 46 and 72 hr after application and scores are recorded. 3. If necrosis is present or the dermal reaction is unusual, the reaction should be described. Severe erythema should receive the maximum score (4), and +N should be used to designate the presence of necrosis and +E the presence of eschar. 4. When the test substance produces dermal irritation that persists 72 hr post-application, daily observations of test and control sites are continued on all animals until all irritation caused by the test substance resolves until Day 14 postapplication.

85

4. The Primary Dermal Irritation Index is calculated for the test substance or control substance by diving the sum of the Total Irritation Scores by the number of observations (three days × three animals = 9 observations). 5. The categories of the Primary Dermal Irritation Index (PDII) are as follows [this categorization of dermal irritation is a modification of the original classification described by Draize et al. (1944)]: PDII = 0.0 >0.0–0.5 >0.5–2.0 >2.0–5.0 >5.0–8.0

nonirritant negligible irritant mild irritant moderate irritant severe irritant

Other abnormalities, such as atonia or desquamation, should be noted and recorded.

D. Evaluation of Results 1. A Subtotal Irritation Value for erythema and eschar formation is determined for each rabbit by adding the values observed at 25, 48, and 72 hr post-application. 2. A Subtotal Irritation Value for edema formation is determined for each rabbit by adding the values observed at 25, 48, and 72 hr postapplication. 3. A Total Irritation Score is calculated for each rabbit by adding the subtotal value for erythema or eschar formation to the subtotal irritation value for edema formation.

Table 8.2

Local Tissue Tolerance

In Vitro Alternatives Extensive progress has been made in devising alternative (in vitro) systems for evaluating the dermal irritation potential of chemicals since this author last reviewed the field (Gad and Chengelis, 1998). Table 8.2 overviews 20 proposed systems which now constitute five very different approaches. This is an effort that extends back to the early 1960s (Choman, 1963).

In Vitro Dermal Irritation Test Systems End-point

Validation data?a

References

Swelling Inhibition of incorporation of [3H]-thymidine and [14C]-leucine labels Leakage of LDH and GOT

No

Dannenberg (1987)

No Yes

Kato et al. (1992) Bartnik et al. (1989)

Morphological evaluation (?)

No

Bell et al. (1989)

Cytotoxicity

No

Naughton et al. (1989)

III. Human epidermal keratinocytes (HEKs) Fibroblasts HEKs HEKs HEKs, dermal fibroblasts HEKs Cultured Chinese hamster ovary Cultured C3 H10T1/2 and HEK cells Cultured cells—BHK21/C13 BHK21/C13 primary rat thymocytes Rat periodontal mast cells

Release of labeled arachidonic acid Acids Cytotoxicity Cytotoxicity (MIT) Cytotoxicity Inflammation mediator release Increases in β-hexosaminidase levels in the media Lipid metabolism inhibition Cell detachment Growth inhibition Increased membrane permeability Inflammation mediator release

Yes Yes Yes Yes No No No

DeLeo et al. (1988) Lamont et al. (1989) Gales et al. (1989) Swisher et al. (1988) Babich et al. (1989) Boyce et al. (1988) Lei et al. (1986) DeLeo et al. (1986)

Yes Yes (surfactants)

Reinhardt et al. (1987) Prottey and Ferguson (1976)

IV. Hen’s egg SKINTEX—protein mixture

Morphological examination Protein coagulation

Yes

Reinhardt et al. (1987) Gordon et al. (1990); Bason et al. 1991

Yes No

Enslein et al.(1987) Firestone and Guy (1986)

System I. Excised patch of perfused skin Mouse skin organ culture Mouse skin organ culture II. TESTSKIN—cultured surrogate skin patch Cultured surrogate skin patch

V. Structure–activity relationship (SAR) model NA SAR model NA

a Evaluated by comparison of predictive accuracy for a range of compounds compared with animal test results. Not validated in the sense used in this chapter. NA = not applicable

86

Safety Evaluation of Medical Devices

The first approaches (I) uses patches of excised human or animal skin maintained in some modification of a glass diffusion cell which maintains the moisture temperature, oxygenation and electrolyte balance of the skin section. In this approach, after the skin section has been allowed to equilibrate for some time, the material of concern is placed on the exterior surface and wetted (if not liquid). Irritation is evaluated either by swelling of the skin (a crude and relatively insensitive method for mild and moderate irritants), by evaluation of inhibition of uptake of radiolabeled nutrients or by measurement of leakage of enzymes through damaged membranes. The second set of approaches (II) utilizes a form of surrogate skin culture comprising a mix of skin cells which closely mirror key aspects of the architecture and function of the intact organ. These systems seemingly offer a real potential advantage but, to date, the “damage markers” employed (or proposed) as predictors of dermal irritation have been limited to cytotoxicity. The third set of approaches (III) is to use some form of cultured cell (either primary of transformed), with primary human epidermal keratinocytes (HEKs) preferred. The cell cultures are exposed to the material of interest, then either ectotoxicity, release of inflammation markers or decrease of some indicator of functionality (lipid metabolism, membrane permeability or cell detachment) is measured. The fourth group (IV) contains two miscellaneous approaches—the use of a membrane from the hen’s egg with morphological evaluation of damage being the predictor of end-point (Reinhardt et al., 1987), and the SKINTEX system, which utilizes the coagulation of a mixture of soluble proteins to predict dermal response. Finally, in group V there are two structure–activity relationship models which use mathematical extensions of past animal results correlated with structure to predict the effects of new structures. Many of these test systems are in the process of evaluation of their performance against various small groups of compounds for which the dermal irritation potential is known. Evaluation by multiple laboratories of a wider range of structures has been performed as a form of validation, particularly under the archives of ECVAM in Europe. As of early 2008, however, there has been no move to adapt any of these methods for use with either medical device or drug safety evaluation.

OCULAR IRRITATION TESTING Ocular irritation is significantly different from the other local tissue irritation tests on a number of grounds (Grant, 1993). For the medical device industry, eye irritation testing is performed when the device is intended to be put into the eye as a means or route of application for ocular therapy. There are a number of special tests applicable to medical devices that are beyond the scope of this chapter, since they are intended to assess potential effects or irritation of a specific device. These are addressed later in the chapter on special cases. In general, however, it is desired that an eye irritation test that is utilized by this group be both sensitive and accurate in predicting the potential to cause irritation in humans. Failing to identify human ocular irritants (lack of sensitivity) is to be avoided, but of equal concern is the occurrence of false positives.

Primary Eye Irritation Test The primary eye irritation test was originally intended to predict the potential for a single splash of chemical into the eye of a human being to cause reversible and/or permanent damage. Since the introduction of the original Draize test 50 years ago (Draize et al., 1944), ocular irritation testing in rabbits has both developed and diverged. Indeed, clearly there is no longer a single test design that is used and different objectives are pursued by different groups using the same test. This lack of standardization has been recognized for some time and attempts have been made to address standardization of at least the methodological aspects of the test, if not the design aspects. One widely used study design, which begins with a screening procedure as an attempt to avoid testing severe irritants or corrosives in animals, goes as follows: A. Test Article Screening Procedure 1. Each test substance will be screened in order to eliminate potentially corrosive or severely irritating materials from being studied for eye irritation in the rabbit. 2. The pH of the test substance (ISO compliant extract or ocular lense solution) measured. 3. A primary dermal irritation test will be performed prior to the study. 4. The test substance will not be studied for eye irritation if it is a strong acid (pH of 2.0 or less) or strong alkali (pH of 11.0 or greater), and/or if the test substance is a severe dermal irritant (with a PDII of 5 to 8) or causes corrosion of the skin. 5. If it is predicted that the test substance does not have the potential to be severely irritating or corrosive to the eye, continue to Section B, Rabbit Screen Procedure. B. Rabbit Screening Procedure 1. A group of at least six New Zealand white rabbits of either sex are screened for the study. The animals are removed from their cages and placed in rabbit restraints. Care should be taken to prevent mechanical damage to the eye during this procedure. 2. All rabbits selected for the study must be in good health; any rabbit exhibiting sniffles, hair loss, loose stools, or apparent weight loss is rejected and replaced. 3. One hour prior to instillation of the test substance, both eyes of each rabbit are examined for signs of irritation and corneal defects with a hand-held slit lamp. All eyes are stained with 2.0% sodium fluorescein and examined to confirm the absence of corneal lesions. Fluorescein Staining: Cup the lower lid of the eye to be tested and instill one drop of a 2% (in water) sodium fluorescein solution onto the surface of the cornea. After 15 sec, thoroughly rinse the eye with physiological saline. Examine the eye, employing a hand-held long-wave ultraviolent illuminator in a darkened room. Corneal lesions, if present, appear as bright yellowish-green fluorescent areas. 4. Only three of the six animals are selected for the study. The three rabbits must not show any signs of eye irritation and must show either a negative or minimum fluorescein reaction (due to normal epithelial desquamation). C. Study Procedure 1. At least 1 hr after fluorescein staining, the test substance is placed in one eye of each animal by gently pulling

Chapter 8:

2. 3. 4. 5.

the lower lid away from the eyeball to form a cup (conjunctival cul-de-sac) into which the test material is dropped. The upper and lower lids are then gently held together for 1 sec to prevent immediate loss of material. The other eye remains untreated (if an ocular lense solution is being tested) or receives just extraction solution, and serves as a control. For testing liquids, 0.01 mL of the test substance is used. The treated eyes of the three rabbits are not washed following instillation of the test substance. To prevent self-inflicted trauma by the animals immediately after instillation of the test substance, the animals are not immediately returned to their cages. After the test and control eyes are examined and graded at 1-hr post-exposure, the animals are returned carefully to their respective cages.

D. Observations 1. The eyes are observed for any immediate signs of discomfort after instilling the test substance. Blepharospasm and/or excessive tearing are indicative of irritating sensations caused by the test substance, and their duration should be noted. Blepharospasm does not necessarily indicate that the eye will show signs of ocular irritation. 2. Grading and scoring of ocular irritation are performed in accordance with Table 8.3. The eyes are examined and grades of ocular reactions are recorded. 3. If signs of irritation persist at Day 7, readings are continued on Days 10 and 14 after exposure or until all signs of reversible toxicity are resolved. 4. In addition to the required observation of the cornea, iris, and conjunctiva, serious effects (such as pannus, rupture of the globe, or blistering of the conjunctivae) indicative of a corrosive action are reported. 5. Whether or not toxic effects are reversible depends on the nature, extent and intensity of damage. Most lesions, if reversible, will heal or clear within 21 days. Therefore, if ocular irritation is present at the 14-day reading, a 21-day reading is required to determine whether the ocular damage is reversible or nonreversible.

Alternatives Testing for potential to cause irritation or damage to the eyes remains the most active area for the development (and validation) of alternatives and the most sensitive area of animal testing in biomedical research. This has been true since the beginning of the 1980s. Table 8.4 presents an overview of the reasons for pursuing such alternatives. The major reason, of course, has been the pressure from public opinion. Indeed, many of the in vitro tests now being evaluated for other end-points (such as skin irritation and lethality) are adaptations of test systems first developed for eye irritation uses. A detailed review of the underlying theory of each test system is beyond the scope of this chapter. Frazier et al. (1987) performed such a review, and Table 8.5 presents an updated version of the list of test systems overviewed in that volume. There are six major categories of approach to replacing in vivo systems (such as the rabbit) and these almost certainly would require some form of battery of such test systems. Many individual systems, however, might constitute effective screens

Table 8.3 Lesionsa

Local Tissue Tolerance

87

Scale of Weighted Scores for Grading the Severity of Ocular

Reaction Criteria I. Cornea A. Opacity degree of density (area that is most dense is taken for reading) 1. Scattered or diffuse area, details of iris clearly visible 2. Easily discernible translucent area, details of iris slightly obscured 3. Opalescent areas, no details of iris visible, size of pupil barely discernible B. Area of cornea involved 1. One-quarter (or less) but not zero 2. Greater than one-quarter, less than one half 3. Greater than one-half, less than whole area 4. Greater than three-quarters up to whole area Scoring equals A × B × 5; total maximum = 80b II. Iris A. Values 1. Folds above normal, congestion, swelling, circumcorneal ingestion (any one or all of these or combination of any thereof), iris still reacting to light (sluggish reaction is possible) 2. No reaction to light, hemorrhage, gross destruction (any one or all of these) Scoring equals A × B (where B is the area of the iris involved, graded as “under cornea”); total maximum = 10 III. Conjunctivae A. Redness (refers to palpebral conjunctivae only) 1. Vessels definitely injected above normal 2. More diffuse, deeper crimson red, individual vessels not easily discernible 3. Diffuse beefy red B. Chemosis 1. Any swelling above normal (include initiating membrane) 2. Obvious swelling with partial eversion of the lids 3. Swelling with lids about half closed 4. Swelling with lids about half closed to completely closed C. Discharge 1. Any amount different from normal (does not include small amount observed in inner canthus of normal animals) 2. Discharge with moistening of the lids and hair just adjacent to the lids 3. Discharge with moistening of the lids and considerable area around the eye Scoring (A + B + C) × 2; total maximum = 20

Score

1 2 3

1 2 3 4

1

2

1 2 3 1 2 3 4 1

2 3

a

The maximum total score is the sum of all scores obtained for the cornea, iris and conjunctivae. b All A × B = Σ(1–3) × Σ(1–4) for three animals.

in defined situations. The first five of these aim at assessing portions of the irritation response, including alterations in tissue morphology, toxicity to individual complete cells or tissue physiology, inflammation or immune modulation, and alterations in repair and/or recovery processes. These methods have the limitation that they assume that one of the component parts can or will predict effects in the complete organ system. While each component may serve well to predict the effects of a set of chemical structures which determine part of the ocular irritation response, a valid assessment across a broad range of structures will require the use of a collection or battery of such tests.

88

Safety Evaluation of Medical Devices

Table 8.4 Rationales for Seeking in vitro Alternatives for Eye Irritancy Tests (1) Avoid whole animal and organ in vivo evaluation. (2) Strict Draize scale testing in the rabbit assesses only three eye structures (conjunctiva, cornea, iris) and traditional rabbit eye irritancy tests do not asses cataracts, pain, discomfort or clouding of the lens. (3) In vivo tests assess only inflammation and immediate structural alternations produced by irritants (not sensitizers, photoirritants or photoallergens). Note, however, that the test was (and generally is) intended to evaluate any pain or discomfort. (4) Technical training and monitoring are critical (particularly in view of the subjective nature of evaluation). (5) Rabbit eye tests do not perfectly predict results in humans, if the objective is either the total exclusion of irritants or the identification of truly severe irritants on an absolute basis (that is, without false positives or negatives). Some (such as Reinhardt et al., 1985) have claimed that these tests are too sensitive for such uses. (6) There are structural and biochemical differences between rabbit and human eyes which make extrapolation from one to the other difficult. For example, Bowman ‘s membrane is present and well developed in man (8-12 µm thick) but not in the rabbit, possibly giving the cornea greater protection. (7) Lack of standardization. (8) Variable correlation with human results. (9) Large biological variability between experimental units. (10) Large, diverse and fragmented databases which are not readily comparable.

The sixth category contains tests that have little or no empirical basis, such as computer-assisted structure– activity relationship models. These approaches can only be assessed in terms of how well or poorly they perform. Table 8.5 presents an overview of all six categories and some of the component tests within them, updated from the assessment by Frazier et al. (1987), along with references for each test. Given that there are no some 70 or more potential in vitro alternatives, the key points along the route to the eventual objective of replacing the in vivo tests systems are thus: (1) How do we select the best candidates from this pool? (2) How do we want to use the resulting system (as a screen or test)? (3) How do we select, develop and validate the system or systems that will actually be used? There have been extensive attempts at validations of many of these tests. Both in the United States and Europe most of the individual investigators have performed such “validations” as part of their development of the test system, and in a number of cases trade associations and groups such as EVCAM and IACVAM have sponsored comparative and/or multi-laboratory validations. At least for screening, several systems should be appropriate for use and, in fact, are used now by several commercial organizations. In terms of use within defined chemical structural classes, use of in vitro systems for testing of chemicals for non-human exposure should supplant traditional in vivo systems once validated on a broad scale by multiple laboratories. Broad use of single tests based on single end-points (such as cytotoxicity) is not likely to be successful, as demonstrated by such efforts as those of Kennah et al. (1989). Again, as of early 2008, there has been no progress towards regulatory acceptance of any of these tests for use with devices or drugs.

Table 8.5

In Vitro Alternatives for Eye Irritation Tests

I. MORPHOLOGY (1) Enucleated superfused rabbit eye system (Burton et al., 1981). (2) Balb/c 3T3 cells/morphological assays (HTD) (Borenfreund and Puerner, 1984). II. CELL TOXICITY (1) Adhesion/cell proliferation (a) BHK cells/growth inhibition (Reinhardt et al., 1985) (b) BHK cells/colony formation efficiency (Reinhardt et al., 1985). (c) BHK cells/cell detachment (Reinhardt et al., (1985). (d) SIRC cells/colony forming assay (North-Root et al., 1982). (e) Balbe/c 3T3 cells/total protein (Shopsis and Eng, 1985). (f) BCL/D1 cells/total protein (Balls and Horner, 1985) (g) Primary rabbit corneal cells/colony forming assay (Watanabe et al., 1988). (2) Membrane integrity (a) LS cells/dual dye staining (Scaife, 1982). (b) Thymocytes/dual fluorescent dye staining (Aeschbacher et al., 1986). (c) LS cells/dual dye staining (Kemp et al., 1983). (d) RCE-SIRC-P815-YAC-1/Cr release (Shadduck et al., 1985). (e) L929 cells/cell variability (Simons, 1981) (f) Bovine red blood cell/hemolysis (Shadduck et al., 1987). (g) Mouse L929 fibroblasts-erythrocin C staining (Frazier, 1988). (h) Rabbit corneal epithelial and endothelial cells/membrane leakage (Meyer and McCulley, 1988). (i) Agarose diffusion (Barnard, 1989). (3) Cell metabolism (a) Rabbit corneal cell cultures/plasminogen activator (Chan, 1985). (b) LS cells/ATP assay (Kemp et al., 1985). (c) Balb/c 3T3 cells/neutral red uptake (Borenfreund and Puerner, 1984)., (d) Balb/c 3T3 uridine uptake inhibition assay (Shopsis and Sathe, 1984). (e) HeLa cells/metabolic inhibition test (MIT-24) (Selling and Ekwall, 1985). (f) MDCK cells/dye diffusion (Tchao, 1988). III. CELL AND TISSUE PHYSIOLOGY (1) Epidermal slice/electrical conductivity (Oliver and Pemberton, 1985). (2) Rabbit ileum/contraction inhibition (Muir et al., 1983). (3) Bovine cornea/corneal opacity (Muir et al., 1984). (4) Proposed mouse eye/permeability test (Maurice and Singh, 1986) IV. INFLAMMATION/IMMUNITY (1) Chorioallantoic membrane (CAM) (a) CAM (Leighton et al., 1983). (b) HET-CAM (Luepke, 1985). (2) Bovine corneal cup model/leukocyte chemotactic factors (Elgebaly et al., 1985). (3) Rat peritoneal mast cells/histamine release (Jacaruso et al., 1985). (4) Rat peritoneal mast cells/serotonin release (Chasin et al., 1979) (5) Rat vaginal explant/prostaglandin release (Dubin et al., 1984). (6) Bovine eye cup/histamine (Hm) and leukotriene C4 (LtC4) release (Benassi et al., 1986). V. RECOVERY/REPAIR (1) Rabbit corneal epithelial cells-wound healing (Jumblatt and Neufeld, 1985). VI. OTHER (1) EYTEX assay (Gordon and Bergman, 1986; Soto et al., 1988) (2) Computer-based structure-activity (SAR) (Enslein, 1984; Enslein et al., 1988).

Chapter 8:

OTHER NONPARENTERAL ROUTE IRRITATION TESTS Mucosal irritation may be evaluated by a number of tests, each of which has serious limitations. In the cheek pouch mucosal test, intact samples or sample extracts are inserted into the cheek pouches of Chinese hamsters. In the vaginal mucosal tests, sample extracts are injected into the vagina of albino rabbits. Rabbits in estrous may give false positive results. In the penile mucosal tests, sample extracts are dripped onto the expressed penises of albino rabbits. Most of the sample is removed when the penis is withdrawn into the body. The oral mucosa and rectal mucosa may also be evaluated. Methods for these tests are set forth in ISO 10993 Part 10. The design of vaginal, rectal, penile, and nasal irritation studies is less formalized, but follows the same basic pattern as the primary dermal irritation test. The rabbit is the preferred species for vaginal and rectal irritation studies, but the monkey and dog have also been used for these (Chvapil, 1979; Eckstein et al., 1969; Lilly et al., 1972; Lindhe et al., 1970; Muller et al., 1988; Nixon et al., 1972; Bernstein and Carlish, 1979; Kaminsky and Willigan, 1982; Davidson, et al. 1982; Haugen, 1980). Both the rabbit and rat have commonly seen use for nasal irritation evaluations. Defined quantities (typically 1.0 mL) of test solutions or suspensions are instilled into the orifice in question. For the vagina or rectum inert bungs are usually installed immediately thereafter to continue exposure for a defined period of time (usually the same period of hours as future human exposure). The orifice is then flushed clean, and 24 hr after exposure it is examined and evaluated (graded) for irritation using the scale in Table 8.1.

PARENTERAL IRRITATION/ TOLERANCE There are a number of special concerns about the safety of materials that are routinely injected (parenterally administered) into the body. By definition, these concerns are all associated with materials that are the products of the pharmaceutical and (in some minor cases) medical device industries. Such parenteral routes include three major ones —IV (intravenous), IM (intramuscular) and SC (subcutaneous)— and a number of minor routes (such as intra-arterial) that are not considered here. These unusual concerns include irritation (vascular, muscular, or subcutaneous), pyrogenicity, blood compatibility, and sterility (Avis, 1985). The background of each of these, along with the underlying mechanisms and factors that influence the level of occurrence of such an effect, are briefly discussed below. Irritation. Tissue irritation upon injection, and the accompanying damage and pain, is a concern that must be addressed for the final formulation, which is to be either tested in humans or marketed, rather than for the active ingredient. This is because most irritation factors are either due to or influenced by aspects of formulation design (see Avis, 1985, for more information on parenteral preparations). These factors are not independent of the route (IV, IM, or SC) that will be used and, in fact (as discussed later) are part of the basis for selecting between the various routes. The lack of irritation and tissue damage at the injection site is sometimes called tolerance. Some of the factors that affect tolerance are not fully under the control of an investigation and are also unrelated to the material being injected. These include body movement, temperature, and animal age.

Local Tissue Tolerance

89

Factors that can be controlled, but are not inherent to the active ingredient, include solubility, tonicity, and pH. And, finally, the active ingredient and vehicle can have inherent irritative effects and factors such as solubility (in the physiological milieu into which they are being injected), concentration, volume, molecular size, and particle size. Gray (1978) and Ballard (1968) discuss these factors and the morphological consequences that may occur if they are not addressed.

PARENTERAL ROUTES There are at least 13 different routes by which to inject material into the body, including the following: 1. Intravenous 2. Subcutaneous 3. Intramuscular 4. Intra-arterial 5. Intradermal 6. Intralesional 7. Epidural

8. Intrathecal 9. Intracisternal 10. Intracardiac 11. Intraventricular 12. Intraocular 13. Intraperitoneal

Each of these routes have devices involved in the administration of the injected or infused agent. Only the first three are discussed in any detail here. Most of these routes of administration place a drug directly or indirectly into systemic circulation. There are a number of these routes, however, by which the drug exerts a local effect, in which case most of the drug does not enter systemic circulation (e.g., intrathecal, intraventricular, intraocular, intracisternal). Certain routes of administration may exert both local and systemic effects depending on the characteristics of the drug and excipients (e.g., subcutaneous). The choice of a particular parenteral route will depend on the required time of onset of action, the required site of action, and the characteristics of the fluid to be injected, among other factors. Muscle irritation is the local inflammation, pain, and damage that result from the parenteral injection of pharmaceuticals into a muscle mass. It is due to a range of physicochemical factors as well as chemical/biological interactions, and is particularly of concern with antibiotics.

Test Systems for Parenteral Irritation There are no regulatory guidelines or suggested test methods for evaluating agents for muscular or vascular irritation. Since such guidelines are lacking, but the evaluation is necessary, those responsible for these evaluations have tried to develop and employ the most scientifically valid procedures. Hagan (1959) first suggested a method for assessing IM irritation, though the need was first identified ten years before (Nelson et al., 1949). His approach, however, did not include a grading system for evaluation of the irritation, and the method used the sacrospinalis muscles, which are somewhat difficult to dissect or repeatedly inject. Shintani et al. (1967) developed and proposed the methodology that currently seems to be more utilized (USP, 2007). It uses the lateral vastus muscle and includes a methodology for evaluation, scoring, and grading of the irritation. Additionally, Shintani et al. (1967) investigated the effects of several factors such as pH of the solution, drug concentration, volume of injection, the effect of repeated injections, and the time to maximum tissue response.

90

Safety Evaluation of Medical Devices

Acute Intramuscular Irritation in the Male Rabbit

Table 8.6

Muscle Irritation Evaluation

Reaction Criteria

1. Overview of Study Design Each rabbit is injected as follows. Site (m. vastus lateralis)

Treatment (1.0 mL/site)

Left Right

(Test article) (Vehicle)

Day 1: Injection of all treatment groups—three rabbits Day 2: Sacrifice and evaluation: 24-hr posttreatment group— three rabbits Day 3: Sacrifice and evaluation: 48-hr posttreatment group— three rabbits Day 4: Sacrifice and evaluation: 72-hr posttreatment group— three rabbits 2. Administration 2.1 Route: The test article is injected into the vastus lateralis of each rabbit. 2.2 Dose: The dose selected is chosen to evaluate the severity of irritation, and represents a concentration that might be used clinically. This volume has been widely used in irritation testing. 2.3 Frequency: Once only. 2.4 Duration: 1 day. 2.5 Volume: 1.0 mL per site.

No discernible gross reaction Slight hyperemia and discoloration Moderate hyperemia and discoloration Distinct discoloration in comparison with the color of the surrounding area Brown degeneration with small necrosis Widespread necrosis with an appearance of “cooked meat” and occasionally an abscess involving the major portions of the muscle

4. In-Life Observations 4.1 Daily observations: once daily following dosing. 4.2 Physical examinations: Once within the 2 weeks before the first dosing day. 4.3 Body weight: Should be determined once before the start of the study. 4.4 Additional examinations may be done by the study director to elucidate any observed clinical signs. 5. Postmortem Procedures 5.1 Irritation is evaluated as follows: Three rabbits are sacrificed by a lethal dose of barbiturate at approximately 24, 48, or 72 hr after dosing. The left and right lateral vastus muscles of each rabbit are excised. The lesions resulting from injections are scored for muscle irritation on a numerical scale of 0 to 5 as in Table 8.6 (Shintani et al., 1967).

0 1 2 3 4 5

The average score for the nine rabbits is then calculated, and a category of irritancy then assigned based on Table 8.7.

Acute Intravenous Irritation in the Male Rabbit The design here is similar to the intramuscular assay, except that injections are made into the veins in specific muscle masses. I. Overview of Study Design Rabbits will be injected as follows:

Group

No. of animals

1

2

2

2

3

2

3. Test System 3.1 Species, age, and weight range: Male New Zealand white rabbits weight 2–5 kg are used. The New Zealand white rabbit has been widely used in muscle irritation research for many years, and is a reasonably sized, even-tempered animal that is well adapted to the laboratory environment. 3.2 Selection: Animals to be used in the study are selected on the basis of acceptable findings from physical examinations and body weights. 3.3 Randomization: Animals are ranked by body weight and assigned a number between 1 and 3. The order of number assigned (e.g., 1–3–2) is chosen from a table of random numbers. Animals assigned number 1 are in the 24-hr post-treatment group, and those assigned number 3 are in the 72-hr post-treatment group.

Score

Treatment site

Evaluation

m. vastus lateralis (left) and cervicodorsal subcutis (left) m. vastus lateralis (right) and cervicodorsal subcutis (right) m. vastus lateralis (left) and cervicodorsal subcutis (left) m. vastus lateralis (right) and cervicodorsal subcutis (right) auricular vein (left) auricular vein (right)

24 hr

72 hr

24- and 72-hr evaluations

Day 1: Injection of all groups (six rabbits) Day 2: Evaluation of Group 3 (two rabbits). Sacrifice and evaluation of Group 1 (two rabbits) Day 4: Evaluation of Group 3 (two rabbits). Sacrifice and evaluation of Group 2 (two rabbits) 2. Administration 2.1. 2.2. 2.3. 2.4.

Intramuscular: M. vastus lateralis Subcutaneous: Cervicodorsal subcutis Intravenous: Auricular vein Dose: The doses and concentration selected are chosen to evaluate the severity of irritation. The dose volumes have been widely used in irritation testing. 2.5. Frequency: Once only. 2.6. Duration: 1 day. 2.7. Volume: M vastus lateralis and cervicodorsal subcutis: 1.0 mL per site; auricular vein: 0.5 mL per site. Table 8.7

Muscle Irritation Categorization

Average Score

Grade

0.0 to 0.4 0.5 to 1.4 1.5 to 2.4 2.5 to 3.4 3.5 to 4.4 4.5 or greater

None Slight Mild Moderate Marked Severe

Chapter 8:

3. Test System 3.1. Species, age, and weight range: Male New Zealand white rabbits, weighing 2–5 kg, are used. 3.2. Selection: Animals to be used in the study are selected on the basis of acceptable findings from physical examinations. 3.3. Randomization: Animals are ranked by body weight and assigned a number between 1 and 3. The order or numbers assigned (e.g., 1–3–2) is chosen from a table of random numbers. Animals assigned number 1 are in Group 1, those assigned number 2 are in Group 2, and those assigned number 3 are in Group 3. 4. In-Life Observations 4.1. Daily observations: Once daily following dosing. 4.2. Physical examinations: Once within the 2 weeks before the first dosing day. 4.3. Body weight: Determined once before the start of the study. 4.4. Additional examinations may be done by the study director to elucidate any observed clinical signs. 5. Postmortem Procedures 5.1. Intramuscular irritation is evaluated as follows: Rabbits are sacrificed by a lethal dose of barbiturate approximately 24 and 72 hr after dosing. The left and right lateral vastus muscles of each rabbit are excised. The reaction resulting from injection is scored for muscle irritation using the scale shown on page 232. 5.2. Subcutaneous irritation is evaluated as follows: Rabbits are sacrificed by a lethal dose of barbiturate approximately 24 and 72 hr after dosing. The subcutaneous injection sites are exposed by dissection, and the reaction is scored for irritation on a scale of 0 to 5, as in Table 8.6. 5.3. Intravenous irritation is evaluated as follows: Rabbits are sacrificed by a lethal dose of barbiturate following the 72-hr irritation evaluation. The injection site and surrounding tissue are grossly evaluated at approximately 24 and 72 hr after dosing on a scale of 0 to 3 as follows: Reaction Criteria

Score

No discernible gross reaction Slight erythema at injection site Moderate erythema and swelling with some discoloration of the vein and surrounding tissue Severe discoloration and swelling of the vein and surrounding tissue with partial or total occlusion of the vein Average score per site Irritancy grade 0.0 to 0.4 None 0.5 to 1.4 Slight 1.5 to 2.4 Moderate 2.5 or greater Severe

0 1 2 3

5.4. Additional examinations may be done by the study director to elucidate the nature of any observed tissue change.

Alternatives Intramuscular (IM) and intravenous (IV) injection of parenteral formulations of pharmaceuticals can produce a range of discomfort, including pain, irritation, and/or damage to muscular or vascular tissue. These are normally evaluated

Local Tissue Tolerance

91

for prospective formulations before use in humans by histopathologic evaluation of damage in intact animal models, usually the rabbit. Attempts have been made to make this in vivo methodology both more objective and quantitative based on measuring the creatinine phosphokinase released in the tissue surrounding the injection site (Meltzer et al., 1970; Sidell et al., 1974). Currently, a protocol utilizing a cultured skeletal muscle cell line (L6) from the rat as a model has been evaluated in an interlaboratory validation program among 11 pharmaceutical laboratories. This methodology (Young et al., 1986) measures creatine kinase levels in media after exposure of the cells to the formulation of interest, and predicts in vivo results for antibiotics and a fair correlation for a broader range of parenteral drug products. Likewise, Williams et al., (1987), Laska et al., (1991) and Kato et al. (1992) have proposed a model that uses cultured primary skeletal muscle fibers from the rat. Damage is evaluated by the release of creatinine phosphokinase. An evaluation using six parenterally administered antibiotics (ranking their EC50 values) showed good relative correlation with in vivo results. Another proposed in vitro assay for muscle irritancy for injectable formulations is the red blood cell hemolysis assay (Brown et al, 1989). Water-soluble formulations in a 1:2 ratio with freshly collected human blood are gently mixed for 5 min. The percentage red blood cell survival is then determined by measuring differential absorbance at 540 mm; this value is then compared to values for known irritants and nonirritants. Against a very small group of compounds (four), this assay reportedly accurately predicts muscle irritation.

INTRACUTANEOUS IRRITATION The Intracutaneous Irritation Test is a sensitive acute toxicity screening test and is generally accepted for detecting potential local irritation by extracts from a biomaterial. Extracts of material obtained with non-irritation polar and nonpolar extraction media are suitable, and sterile extracts are desirable.

Intracutaneous Test This test is designed to evaluate local responses to the extracts of materials under test following intracutaneous injection into rabbits.

Test Animal Select healthy, thin-skinned albino rabbits whose fur can be clipped closely and whose skin is free from mechanical irritation or trauma. In handling the animals, avoid touching the injection sites during observation periods, except to discriminate between edema and an oil residue. Rabbits previously used in unrelated tests, such as the Pyrogen Test, and that have received the prescribed rest period, may be used for this test provided that they have clean unblemished skin.

Procedure Agitate each extract vigorously prior to withdrawal of injection doses to ensure even distribution of the extracted matter. On the day of the test, closely clip the fur on the animal’s back on both sides of the spinal column over a sufficiently large test area. Avoid mechanical irritation and trauma. Remove loose hair by means of vacuum. If necessary, swab

92

Safety Evaluation of Medical Devices

Table 8.8

Intracutaneous Test

Extract or Blank Sample Blank

Number of sites(per animal)

Dose, µL per site

5 5

200 200

the skin lightly with diluted alcohol, and dry the skin prior to injection. More than one extract from a given material can be used per rabbit, if you have determined that the test results will not be affected. For each Sample use two animals and inject each intracutaneously, using one side of the animal for the Sample and the other side for the Blank, as outlined in Table 8.8. [Note—Dilute each gram of the extract of the Sample prepared with Polyethylene Glycol 400, and the corresponding Blank, with 7.5 volumes of Sodium Chloride Injection to obtain a solution having a concentration of about 120 mg of polyethylene glycol per mL.] Examine injection sites for evidence of any tissue reaction such as erythema, edema, and necrosis. Swab the skin lightly, if necessary, with diluted alcohol to facilitate reading of injection sites. Observe all animals at 24, 48, and 72 hours after injection. Rate the observations on a numerical scale for the extract of the Sample and for the Blank, using Table 8.1. Reclip the fur as necessary during the observation period. If each animal at any observation period shows an average reaction to the Sample that is not significantly greater than to the Blank, the Sample meets the requirements of this test. If at any observation period the average reaction to the Sample is questionably greater than the average reaction to the Blank, repeat the test using three additional rabbits. On the repeat test, the average reaction to the Sample in any of the three animals is not significantly greater than the Blank.

Pyrogenicity Pyrogenicity is the induction of a febrile (fever) response by the parenteral (usually IV or IM) administration of exogenous material, usually (but not always) bacterial endotoxins. Pyrogenicity is usually associated with microbiological contamination of a final formulation or product but is now increasingly of concern because of the increase in interest in biosynthetically produced materials. Generally, ensuring the sterility of product and process will guard against pyrogenicity. Pyrogenicity testing is performed extensively in the medical device industry. If a device is to be introduced directly or indirectly into the fluid path, it is required that it be evaluated for pyrogenic potential (European Pharmacopoeia, 1990; USP, 2007). The USP Pyrogen Test using rabbit or the Limulus Amebocyte Lysate (LAL) test can be used to support pyrogen free claims. If the LAL test is used, the LAL test method must either meet the FDA’s documented titled Guideline on Validation of the Limulus Amebocyte Lysate Test as an EndProduct Endotoxin Test for Human and Animal Parental Drugs, Biological Products, and Medical Devices or a 510(k) or PMA application must be submitted for the LAL test. The bacterial endotoxin limit for medical devices is 0.5 EU/mL. Manufacturers may retest LAL test failures with another LAL test or the USP rabbit pyrogen test. Medical devices that contact cerebrospinal fluid should have less than 0.06 EU/mL of endotoxin. In vitro pyrogenicity testing (or bacterial endotoxin testing) is one of the great success stories for in vitro testing. Some 30 years ago, the limulus amebocyte lysate (LAL) test was developed, validated, and accepted as an in vitro test for

estimating the concentration of bacterial endotoxins that may be present in or on a sample of the article(s) to which the test is applied uses LAL that has been obtained from aqueous extracts of the circulating amebocytes of the horseshoe crab, Limulus polyphemus and that has been prepared and characterized for use as an LAL reagent for gel-clot formation (Cooper, 1975; Weary and Baker, 1977). The test’s limitation is that it detects only the pyrogens of gram-negative bacteria. This is generally not significant (at least for use in lot release assays) since most environmental contaminants that gain entrance to sterile products are gram-negative (Bulich et al., 1981; Devleeschouwer et al., 1985). Where the test is conducted as a limit test, the specimen is determined to be positive or negative to the test judged against the endotoxin concentration specified in the individual monograph (USP, 2007). Where the test is conducted as an assay of the concentration of endotoxin, with calculation of confidence limits of the result obtained, the specimen is judged to comply with the requirements if the result does not exceed (1) the concentration limit specified in the individual monograph and (2) the specified confidence limits for the assay. In either case the determination of the reaction end point is made with parallel dilutions of redefined endotoxin units. Since LAL reagents have also been formulated to be used for turbidimetric (including kinetic) assays or colorimetric readings, such tests may be used if shown to comply with the requirements for alternative methods. These tests require the establishment of a standard regression curve and the endotoxin content of the test material is determined by interpolation from the curve. The procedures include incubation for a preselected time of reacting endotoxin and control solutions with LAL reagent and reading the spectrophotometric light absorbance at suitable wavelengths. In the case of the turbidimetric procedure the reading is made immediately at the end of the incubation period. In the kinetic assays, the absorbance is measured throughout the reaction period and rate values are determined from those readings. In the colorimetric procedure the reaction is arrested at the end of the preselected time by the addition of an appropriate amount of acetic acid solution prior to the readings. A possible advantage in the mathematical treatment of results, if the test is otherwise validated and the assay suitable designed, could be the confidence interval and limits of potency from the internal evidence of each assay itself.

Reference Standard and Control Standard Endotoxins The reference standard endotoxin (RSE) is the USP Endotoxin Reference Standard, which has a defined potency of 10,000 USP endotoxin units (EU) per vial. Constitute the entire contents of one vial of the RSE with 5 mL of LAL reagent water, vortex for not less than 20 min and use this concentrate for making appropriate serial dilutions. Preserve the concentrate in a refrigerator, for making subsequent dilutions, for not more than 14 days. Allow it to reach room temperature, if applicable, and vortex it vigorously for not less than 5 min before use. Vortex each dilution for not less than 1 min before proceeding to make the next dilution. Do not use stored dilutions. A control standard endotoxin (CSE) is an endotoxin preparation other than the RSE that has been standardized against the RSE. If a CSE is a preparation not already adequately characterized, its evaluation should include characterizing parameters both for endotoxin

Chapter 8:

quality and performance (such as reaction in the rabbit), and for suitability of the material to serve as a reference (such as uniformity and stability). Detailed procedures for its weighing and/or constitution and use to ensure consistency in performance should also be included. Standardization of CSE against the RSE using an LAL reagent for the gel-clot procedure may be effected by assaying a minimum of four vials of the CSE or four corresponding aliquots, where applicable, of the bulk CSE and one vial of the RSE as directed under Test Procedure, but using four replicate reaction tubes at each level of the dilution series for the RSE and four replicate reaction tubes similarly for each vial or aliquot of the CSE. If the dilutions for the four vials or aliquots of the CSE cannot all be accommodated with the dilutions for the one vial of the RSE on the same rack for incubation, additional racks may be used for accommodating some of the replicate dilutions for the CSE, but all of the racks containing the dilutions of the RSE and CSE are incubated as a block. However, in such cases, the replicate dilution series from the one vial of the RSE are accommodated together on a single rack and the replicate dilution series from any one of the four vials or aliquots of the CSE are not divided between racks. The antilog of the difference between the mean log 10 end point of the RSE and the mean log 10 end point of the CSE is the standardized potency of the CSE, which is then converted to and expressed in units/ng under stated drying conditions for the CSE, or units per container, whichever is appropriate. Standardize each new lot of CSE prior to use in the test. Calibration of a CSE in terms of the RSE must be with the specific lot of LAL reagent and the test procedure with which it is to be used. Subsequent lots of LAL reagent from the same source and with similar characteristics need only checking of the potency ratio. The inclusion of one or more dilution series made from the RSE when the CSE is used for testing will enable observation of whether or not the relative potency shown by the latter remains within the determined confidence limits. A large lot of a CSE may, however, be characterized by a collaborative assay of a suitable design to provide a representative relative potency and the within-laboratory and between-laboratory variance. A suitable CSE has a potency of not less than 2 EU/ng and not more than 50 EN/ng, where in bulk form, under adopted uniform drying conditions, e.g., to a particular low moisture content and other specified conditions of use, and a potency within a corresponding range where filled in vials of a homogeneous lot.

Preparatory Testing Use an LAL agent of confirmed label or determined sensitivity. In addition, where there is to be a change in lot of CSE, LAL reagent, or another reagent, conduct tests of a prior satisfactory lot of CSE, LAL, and/or other reagent in parallel on changeover. Treat any containers or utensils employed so as to destroy extraneous surface endotoxins that may be present, such as by heating in an oven at 250ºF or above for sufficient time. The validity of test results for bacterial endotoxins requires an adequate demonstration that specimens of the article, or of solutions, washings, or extracts thereof to which the test is to be applied, do not of themselves inhibit or enhance the reaction or otherwise interfere with the test. Validation is accomplished by testing untreated specimens or appropriate dilutions thereof, concomitantly with and without known and demonstrable added amounts of RSE or a CSE, and comparing the results obtained. Appropriate negative

Local Tissue Tolerance

93

controls are included. Validation must be repeated if the LAL reagent source or the method of manufacture or formulations of the article is changed.

Test for Confirmation of Labeled LAL Reagent Sensitivity Confirm the labeled sensitivity of the particular LAL reagent with the RSE (or CSE) using not less than four replicate vials, under conditions shown to achieve an acceptable variability of the test, viz., the antilog of the geometric mean log 10 lystate gel-clot sensitivity is within 0.5 to 2.0, where the labeled sensitivity is in EU/mL. The RSE (or CSE) concentrations selected to confirm to LAL reagent label potency should bracket the stated sensitivity of the LAL reagent. Confirm the labeled sensitivity of each new lot of LAL reagent prior to use in the test.

Inhibitions or Enhancement Test Conducts assays, with standard endotoxin, or untreated specimens in which there is no endogenous endotoxin detectable, and of the same specimens to which endotoxin has been added, as directed under Test Procedures, but use not less than four replicate reaction tubes at each level of the dilution series for each untreated specimen and for each specimen to which endotoxin has been added. Record the end points (E, in units/mL) observed in the replicates. Take the logarithms (e) of the end points, and compute the geometric means of the log end points for the RSE (or CSE) for the untreated specimens and for specimens containing endotoxin by the formula antilog, elf, where e is the sum of the log end points of the dilution series used and f is the number of replicate end points in each case. Compute the amount of endotoxin in the specimen to which endotoxin has been added. The test is valid for the article if this result is within twofold of the known added amount of endotoxin. Alternatively, if the test has been appropriately set up, the test is valid for the article if the geometric mean end-point dilution for the specimen to which endotoxin has been added is within one twofold dilution of the corresponding geometric mean end-point dilution of the standard endotoxin. Repeat the test for inhibition or enhancement using specimens diluted by a factor not exceeding that given by the formula, x/y (see Maximum Valid Dilution, below). Use the least dilution sufficient to overcome the inhibition or enhancement of the known endotoxin for subsequent assays of endotoxin in test specimens. If endogenous endotoxin is detectable in the untreated specimens under the conditions of the test, the article is unsuitable for the inhibition or enhancement test, or it may be rendered suitable by removing the endotoxin present by untrafiltration or by appropriate dilution. Dilute the untreated specimen (as constituted, where applicable, for administration or use) to a level not exceeding the maximum valid dilution, at which no endotoxin is detectable. Repeat the test for inhibition or enhancement using the specimens at those dilutions.

Test Procedure In preparing for and applying the test, observe precautions in handling the specimens in order to avoid gross microbial contamination. Washings or rinsings of devices must be with LAL reagent water in volumes appropriate to their use and, where applicable, of the surface area which comes into contact with body tissues or fluids. Use such washings or rinsings if the extracting fluid has been in contact with the relevant pathway or surface for not less than 1 hr at controlled

94

Safety Evaluation of Medical Devices

room temperature (15–30ºC). Such extracts may be combined, where appropriate. For validating the test for an article, for endotoxin limit tests or assays, or for special purposes where so specified, testing of specimens is conducted quantitatively to determine response end points for gel-clot readings. Usually graded strengths of the specimen and standard endotoxin are made by multifold dilutions. Select dilutions so that they correspond to a geometric series in which each step is greater than the next lower by a constant ratio. Do not store diluted endotoxin, because of loss of activity by absorption. In the absence of supporting data to the contrary, negative and positive controls are incorporated into the test. Use not less than two replicate reaction tubes at each level of the dilution series for each specimen under test. Whether the test is employed as a limit test or as a quantitative assay, a standard endotoxin dilution series involving not less then two replicate reaction tubes is conducted in parallel. A set of standard endotoxin dilution series is included for each block of tubes, which may consist of a number of racks for incubation together, provided the environmental conditions within blocks are uniform.

Preparation Since the form and amount per container of standard endotoxin and of LAL reagent may vary, constitution and/or dilution of contents should be as directed in the labeling. The pH of the test mixture of the specimen and the LAL reagent is in the range of 6.0–7.5 unless specifically directed otherwise in the individual monograph. The pH may be adjusted by the addition of sterile, endotoxin-free sodium hydroxide or hydrochloric acid or suitable buffers to the specimen prior to testing.

Maximum Valid Dilution The maximum valid dilution (MVD) is appropriate to injections or to solutions for parenteral administration in the form constituted or diluted for administration, or, where applicable, to the amount of drug by weight if the volume of the dosage form for administration could be varied. Where the endotoxin limit concentration is specified in the individual monograph in terms of volume (in EU/mL), divide the limit by γ, which is the labeled sensitivity (in EU/mL) of the lysate employed in the assay, to obtain the MVD factor. Where the endotoxin limit concentration is specified in the individual monograph in terms of weight or of units of active drug (in EU/mg or in EU/unit), multiply the limit by the concentration (in mg/mL or in units/mL of the drug in the solution tested or of the drug constituted according to the label instructions, whichever is applicable, and divide the product of the multiplication by γ to obtain the MVD factor. The MVD factor so obtained is the limit dilution factor for the preparation for the test to be valid.

Procedure

To 10 × 75-mm test tubes add aliquots of the appropriately constituted LAL reagent and the specified volumes of specimens, endotoxin standard, negative controls, and a positive product control consisting of the article, or of solutions, washings, or extracts thereof, to which the RSE (or a standardized CSE) has been added at a concentration of endotoxin of 2 for LAL reagent (see under Test for confirmation of labeled LAL reagent sensitivity). Swirl each gently to mix and place in an incubating device such as water bath or heating block, accurately recording the time at which the

tubes are so placed. Incubate each tube, undisturbed, for 60 ±2 min at 37± 1ºC, and carefully remove it for observation. A positive reaction is characterized by the formation of a firm gel that remains when inverted through 180 degrees. Record such a result as a positive (+). A negative result is characterized by the absence of such a gel or by the formation of a fiscous gel that does not maintain its integrity. Record such a result as a negative (−). Handle the tubes with care, and avoid subjecting them to unwanted vibrations, or false-negative observations may result. The test is invalid if the positive product control or the endotoxin standard does not show the end-point concentration to be within ± twofold dilutions from the label claim sensitivity of the LAL reagent or if any negative control shows a gel-clot end point.

Calculation and Interpretation Calculate the concentration of endotoxin (in units/mL or in units/g or mg) in or on the article under test by the formula pS/U, where S is the antilog of the geometric mean log 10 of the end points, expressed in EU/mL for the standard endotoxin, U is the antilog of elf, where e is the log 10 of the end-point dilution factors, expressed in decimal fractions, f is the number of replicate reaction tubes read at the end-point level for the specimen under test, and p is the correction factor for those cases where a specimen of the article cannot be taken directly into test but is processed as an extract, solution, or washing. Where the test is conducted as an assay with sufficient replication to provide a suitable number of independent results, calculate for each replicate assay the concentration of endotoxin in or on the article under test from the antilog of the geometric mean log end-point ratios. Calculate the mean and the confidence limits from the replicate logarithmic values of all the obtained assay results by a suitable statistical method.

Interpretation The article meets the requirements of the test if the concentration of endotoxin does not exceed that specified in the individual monograph, and the confidence limits of the assay do not exceed those specified.

Rabbit Pyrogen Test The United States Pharmacopeia describes a pryogen test using rabbits as a model. This test, which is the standard for limiting risks of a febrile reaction to an acceptable level, involves measuring the rise in body temperature in a group of three rabbits for 3 hr after injection of 10 mL of test solution.

Apparatus and Diluents Render the syringes, needles, and glassware free of pyrogens by heating at 250ºF for not less than 30 min or by any other suitable method. Treat all diluents and solutions by washing and rinsing of devices or parenteral injection assemblies in a manner that will ensure that they are sterile and pyrogenfree. Periodically perform control pyrogen tests on representative portions of the diluents and solutions that are used for washing and rinsing of the apparatus.

Temperature Recording Use an accurate temperature-sensing device, such as a clinical thermometer or thermistor or similar probe, that has been calibrated to ensure an accuracy of 0.1 and has been tested to determine that a maximum reading is reached in less than 5 min. Insert the temperature-sensing probe into the rectum of

Chapter 8:

the test rabbit to a depth of not less than 7.5 cm and, after a period of time not less than that previously determined as sufficient, record the rabbit’s temperature.

Test Animals Use healthy, mature rabbits. House the rabbits individually in an area of uniform temperature (between 20 and 23ºC) free from disturbances likely to excite them. The temperature should vary no more than ±3ºC from the selected temperature. Before using a rabbit for the first time in a pyrogen test, condition it for not more than seven days before use by a sham test that includes all of the steps as directed under Procedure, except injection. Do not use a rabbit for pyrogen testing more frequently than once every 48 hr, nor prior to two weeks following a maximum rise in its temperature of 0.6º or more while being subjected to the pyrogen test, or following its having been given a test specimen that was found to be pyrogenic.

Procedure Perform the test in a separate area designated solely for pyrogen testing and under environmental conditions similar to those under which the animals are housed. Withhold all food from the test rabbits during the period of the test. Access to water is allowed at all times, but may be restricted during the test. If probes measuring rectal temperature remain inserted throughout the testing period, restrain the rabbits with loose-fitting Elizabethan collars that allow the rabbits to assume a natural resting posture. Not more than 30 min prior to the injection of the test dose, determine the “control temperature” of each rabbit; this is the base for the determination of any temperature increase resulting from the injection of a test solution. In any one group of test rabbits, use only those rabbits whose control temperatures do not vary by more than 1ºC from each other, and do not use any rabbit having a temperature exceeding 39.8ºC. Unless otherwise specified in the individual protocol, inject 10 mL of the test solution per kilogram of body weight into an ear vein of each of three rabbits, completing each injection within 10 min after the start of administration. The test solution is either the product, constituted if necessary as directed in the labeling, or the material under test. For pyrogen testing of devices or injection assemblies, use washings or rinsings of the surfaces that come in contact with the parenterally administered material or with the injection site or internal tissues of the patient. Ensure that all test solutions are protected from contamination. Perform the injection after warming the test solution to a temperature of 37ºC ± 2º. Record the temperature at 1, 2, and 3 hr subsequent to the injection.

Test Interpretation and Continuation Consider any temperature decreases as zero rise. If no rabbit shows an individual rise in temperature of 0.6º or more above its respective control temperature, and if the sum of the three individual maximum temperature rises does not exceed 1.4º, the product meets the requirements for the absence of pryogens. If any rabbit shows an individual temperature rise of 0.6º or more, or if the sum of the three individual maximum temperature rises exceeds 1.4º, continue the test using five other rabbits. If not more than three of the eight rabbits show individual rises in temperature of 0.6º or more, and if the sum of the eight individual maximum temperature rises does not exceed 3.7º, the material under examination meets the requirements for the absence of pyrogens.

Local Tissue Tolerance

95

FACTORS AFFECTING IRRITATION RESPONSES AND TEST OUTCOME The results of local tissue irritation tests are subject to considerabale variability due to relatively small differences in test design or technique. Weil and Scala (1971) arranged and reported on the best known of several intra-laboratory studies to clearly establish this fact. Though the methods presented above have proven to give reproducible results in the hands of the same technicians over a period of years (Gad et al., 1986) and contain some internal controls (the positive and vehicle controls in the PDI) against large variabilities in results or the occurrence of either false positives or negatives, it is still essential to be aware of those factors that may systematically alter test results. These factors are summarized below. In general, any factor that increases absorption through the stratum corneum or mucous membrane will also increase the severity of an intrinsic response. Unless this factor mirrors potential exposure conditions, it may, in turn, adversely affect the relevance of test results. The physical nature of solids must be carefully considered both before testing and in interpreting results. Shape (sharp edges), size (small particles may abrade the skin due to being rubbed back and forth under the occlusive wrap), and rigidity (stiff fibers or very hard particles will be physically irritating) of solids may all enhance an irritation response. Solids frequently give different results when they are tested dry than if wetted for a test. As a general rule, solids are more irritating if moistened (going back to Item A, wetting is a factor that tends to enhance absorption). Care should also be taken as to moistening agent—some (few) batches of U.S. Pharmacopeia physiological saline have proven to be mildly irritating to the skin and mucous membrane on their own. Liquids other than water or saline should not be used. If the treated region on potential patients will be a compromised skin surface barrier (e.g., if it is cut or burned), some test animals should likewise have their application sites compromised. This procedure is based on the assumption that abraded skin is uniformly more sensitive to irritation. Experiments, however, have shown that this is not necessarily true; some materials produce more irritation on abraded skin, while others produce less (Guillot et al., 1982; Gad et al., 1986). The degree of occlusion (in fact, the tightness of the wrap over the test site) also alters percutaneous absorption and therefore irritation. One important quality control issue in the laboratory is achieving a reproducible degree of occlusion in dermal wrappings. Both the age of the test animal and the application site (saddle of the back vs. flank) can markedly alter test outcome. Both of these factors are also operative in humans, of course (Mathias, 1983), but in dermal irritation tests, the objective is to remove all such sources of variability. In general, as an animal ages, sensitivity to irritation decreases. For the dermal test, the skin on the middle of the back (other than directly over the spine) tends to be thicker (and therefore less sensitive to irritations) than that on the flanks. The sex of the test animals can also alter study results, because both regional skin thickness and surface blood flow vary between males and females. Finally, the single most important (yet also most frequently overlooked) factor that influences the results and outcome of these (and, in fact, most) acute studies is the

96

Safety Evaluation of Medical Devices

training of the staff. In determining how test materials are prepared and applied and in how results are “read” against a subjective scale, both accuracy and precision are extremely dependent on the technicians involved. To achieve the desired results, initial training must be careful and all-inclusive. As important, some form of regular refresher training must be exercised—particularly in the area of scoring of results. Use of a set of color photographic standards as a training and reference tool is strongly recommended; such standards should clearly demonstrate each of the grades in the Draize dermal scale. It should be recognized that local tissue tolerance (or “irritancy”) tests are designed with a bias to preclude false negatives and, therefore, tend to exaggerate results in relation to what would happen in humans. Findings of negligible irritancy (or even in the very low mild irritant range) should therefore be of no concern unless the product under test is to have large-scale and prolonged dermal contact.

REFERENCES Aeschbacher, M., Reinhardt, C.A. and Zbinden, G. (1986). A rapid cell membrane permeability test using fluorescent dyes and flow cytometry. Cell. Biol. Toxicol., 2: 247. ASTM (1981). Standard Practice for Testing Biomaterials in Rabbits for Primary Skin Irritation, F719-81. Avis, K.E. (1985). Parenteral preparations. In: A.R. Gennaro (Ed.), Remington’s Pharmaceutical Sciences. Mack Publishing Company, Easton, Pennsylvania, pp. 1518–1541. Babich, H., Martin-Alguacil, N. and Borenfreund, E. (1989). Comparisons of the cytotoxicities of dermatotoxicants to human keratinocytes and fibroblasts in vitro. In: A.M. Goldberg (Ed.), In Vitro Toxicology: New Directions. Mary Ann Liebert, New York, pp. 153–167. Ballard, B.E. (1968). Biopharmaceutical considerations in subcutaneous and intramuscular drug administration. J. Pharm. Sci., 57: 357–378. Balls, M. and Horner, S.A. (1985). The FRAME interlaboratory program on in vitro cytotoxicology. Food Chem. Toxic., 23: 205–213. Barnard, N.D. (1989). A Draize alternative. The Animal’s Agenda, 6: 45. Bartnik, F.G., Pittermann, W.F., Mendoft, N., Tillmann, U. and Kunstler, K. (1989). Skin organ culture for the study of skin irritancy. Third International Congress of Toxicology. Brighton, U.K. Bason, M., Harvell, J., Gordon, V., Maibach, H. (1991). Evaluation of the SKINTEX system. Presented at the Irritant Contact Dermititis Symposium. Int. J. Dermatol., 30: 623–626. Groningen, Netherlands. Bell, E., Gay R. and Swiderek, M. (1989). Use of fabricated living tissue and organ equivalents as defined in higher order systems for the study of pharmacologic responses to test substances. Presented at the NATO Advanced Research Workshop. Pharmaceutical Application of Cell and Tissue Culture to Drug Transport. Bandol, France, September 4–9, 1989. Benassi, C.A., Angi, M.R., Salvalaoi, L. and Bettero, A. (1986). Histamine and leukotriene C4 release from isolated bovine sclerachoroid complex: A new in vitro ocular irritation test. Chim. Agg., 16: 631–634. Bernstein, M.L. and Carlish, K. (1979). The induction of hyperkeratotic white lesions in hamster cheek pouches with mouthwash. Oral Surg., 48: 517. Borenfreund, E. and Puerner, J.A. (1984). A simple quantitative procedure using monolayer cultures for cytotoxicity assays. J. Tissue Culture Methods, 9: 7–9. Borenfreund, E. and Peurner, J.A. (1985). Toxicity determined in vitro by morphological alterations and Neutral Red absorption. Toxicol. Lett., 24: 119–124. Boyce, S.T., Hansbrough, J.F. and Norris, D.A. (1988). Cellular responses of cultured human epidermal keratinocytes as models

of toxicity to human skin. In: A.M. Goldberg (Ed.), Progress in In Vitro Toxicology. Mary Ann Liebert, New York, pp. 23–37. Brown, S., Templeton, D. Prater, A. and Potter, C. J. (1989). Use of an in vitro hemolysis test to predict tissue irritancy in an intramuscular formulation. J. Parenter. Sci. Technol., 43: 117–120. Bulich, A.A., Greene, M.W. and Isenberg, D.L. (1981). Reliability of bacterial compounds and complex effluents. In: D.R. Branson, K.L. Dickson (Eds.), Aquatic Toxicology and Hazard Assessment. American Society for Testing and Materials (ASTM 737), Philadelphia, 338–347. Burton, A.B.G., York, M. and Lawrence, R.S. (1989). The in vitro assessment of severe eye irritation. Food Cosmet. Toxicol., 19: 471–480. Chan, K.Y. (1985). An in vitro alternative to the Draize test. In: A.M. Goldberg (Ed.), In Vitro Toxicology: Alternative Methods in Toxicology, Vol. 3. Mary Ann Liebert, New York, pp. 405–422. Chasin, M., Scott, C., Shaw, C. and Persico, F. (1979). A new assay for the measurement of mediator release from rat peritoneal mast cells. Int. Arch. Allergy Appl. Immunol., 58: 1–10. Choman, B.R. (1963). Determination of the response of skin to chemical agents by an in vitro procedure. J. Invest. Dermatol., 44: 177–182. Chvapil, M., Chvapil, T.A., Owen, J.A., Kantor, M., Ulreich, J.B. and Eskelson, C. (1979). Reaction of vaginal tissue of rabbits to inserted sponges made of various materials. J. Biomed. Mater. Res., 13: 1. Cooper, J. F. (1975). Principles and applications of the limulus test for pyrogen in parenteral drugs. Bull. Parenter. Drug Assoc., 3: 122–130. Dannenberg, A.M., Moore, K.G., Schofield, B.H. et al. (1987). Two new in vitro methods for evaluating toxicity in skin (employing short-term organ culture). In: A.M. Goldberg (Ed.), Alternative Methods in Toxicology, Vol. 5. Mary Ann Liebert, New York, pp. 115–128. Davidson, W.M., Sheinis, E.M. and Shepherd, S.R. (1982). Tissue reaction to orthodontic adhesives. Am. J. Orthod., 82: 502. DeLeo, V., Hong, J., Scheide, S. Kong, B., DeSalva, S. and Bagley, D. (1988). Surfactant-induced cutaneous primary irritancy: An in vitro model-assay system development. In: A.M. Goldberg (Ed.), Progress in In Vitro Technology. Mary Ann Liebert, New York, pp. 39–43. DeLeo, V., Midlarsky, L., Harber, L.C., Kong, B.M. and DeSalva, S. (1987). Surfactant-induced cutaneous primary irritancy: An in vitro model. In: A.M. Goldberg (Ed.), Alternative Methods in Toxicology, Vol. 5. Mary Ann Liebert, New York, pp. 129–138. Devleeschouwer, M.J., Cornil, M. F. and Dony, J. (1985). Studies on the sensitivity and specificity of the limulus amebocyte lysate test and rabbit pyrogen assays. Appl. Environ. Micriobiol., 50: 1509–1511. Draize, J.H. (1955). Dermal Toxicity. Association of Food and Drug Officials of the U.S., pp. 46–59. FDA, Washington, D.C. Draize, J.H. (1959). Dermal toxicity. Appraisal of the Safety of Chemicals in Foods, Drugs, and Cosmetics. Austin, Texas, Association of Food and Drug Officials of the United States, Texas State Department of Health, Texas. Draize, J. H., Woodard, G. and Clavery, H. O. (1944). Method for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. J. Pharmacol. Exp. Ther., 82: 237–390. Dubin, N.H., De Blasi, M.C. et al. (1984). Development of an in vitro test for cytotoxicity in vaginal tissue: effect of ethanol on prostanoind release. In: A.M. Goldberg (Ed.), Acute Toxicity Testing: Alternative Approaches. Alternative Methods in Toxicology, Vol. 2. Mary Ann Liebert, New York, pp. 127–138. Eckstein, P., Jackson, M.C.N., Millman, N. and Sobrero, A.J. (1969). Comparison of vaginal tolerance tests of spermicidal preparations in rabbits and monkeys. J. Reprod. Fert., 20: 85–93. Elgebaly, S.S., Nabawi, K., Herkbert, N., O’Rourke, J. and Kruetzer, D.L. (1985). Characterization of neutrophil and monocyte specific chemotactic factors derived from the cornea in response to injury. Invest. Ophthalmol. Vis. Sci., 26: 320. Enslein, K. (1984). Estimation of toxicology end points by structureactivity relationships. Pharmacol. Rev., 36: 131–134.

Chapter 8: Enslein, K., Blake, V.W., Tuzzeo, T.M., Borgstedt, H.H., Hart, J.B. and Salem, H. (1988). Estimation of rabbit eye irritation scores by structure-activity relationships. In Vitro Toxicol., 2: 1–14. Enslein, K., Borgstedt, H.H., Blake, B.W. and Hart, J.B. (1987). Prediction of rabbit skin irritation severity by structure-activity relationships. In Vitro Toxicol., 1: 129–147. European Pharmacopoeia (1990). Part V.2.1.9. Pyrogens. Firestone, B.A. and Guy, R.H. (1986). Approaches to the prediction of dermal absorption and potential cutaneous toxicity. In: A.M. Goldberg (Ed.), In Vitro Toxicology. Alternative Methods in Toxicology, Vol. 3. Mary Ann Liebert, New York, pp. 516–536. Frazier, J.M. (1988). Update: A critical evaluation of alternatives to acute ocular irritancy testing. In: A.M. Goldberg (Ed.), Progress in In Vitro Toxiciology. Mary Ann Liebert, New York, pp. 67–75. Frazier, J.M., Gad, S.C., Goldberg, A.M. and McCulley, J.P. (1987). A Critical Evaluation of Alternatives to Acute Ocular Irritancy Testing. Mary Ann Liebert, New York. Gad. S.C. and Chengelis, C.P. (1998). Acute Toxicity: Principles and Methods, Second Edition. Academic Press, San Diego, CA. Gad, S.C., Walsh, R.D. and Dunn B.J. (1986). Correlation of ocular and dermal irritancy of industrial chemicals. Ocular and Dermal Toxicology, 5(3): 195–213. Gales, Y.A., Gross, C.L., Karebs, R.C. and Smith, W.J. (1989). Flow cytometric analysis of toxicity by alkylating agents in human epidermal keratinocytes. In: A.M. Goldberg (Ed.), In Vitro Toxicology: New Directions. Mary Ann Liebert, New York, pp. 169–174. Geller, W., Kobel, W. and Seifert, G. (1985). Overview of animal test methods for skin irritation. Fundam. Chem. Toxicol., 23(2): 165–168. Gordon, V.C. and Bergman, H.C. (1986). Eyetex, an In Vitro Method for Evaluation of Optical Irritancy, Report, 26. National Testing Corporation. Gordon, V.C., Kelly, C.P. and Bergman, H.C. (1990). Evaluation of SKINTEX, an in vitro method for determining dermal irritation. Toxicologist, 10(1): 78. Grant, M.W. (1993) Toxicology of the Eye. 4th Edn. Thomas, Springfield, Ill. Gray, J.E. (1978). Pathological evaluation of infection injury. In: Robinson, J. (Ed.), Sustained and Controlled Release Drug Delivery Systems. Marcel Dekker, New York, pp. 351–405. Guillot, J.P., Gonnet, J.F., Clement, C., Caillard, L. and Trahaut, R. (1982). Evaluation of the cutaneous-irritation potential of 56 compounds. Fd. Chem. Toxicol., 201: 563–572. Hagan, E.C. (1959). Appraisal of the Safety of Chemicals in Foods, Drugs and Cosmetics. Association of Food and Drug Officials of the United States, Austin, Texas, p. 19. Haugen, E. (1980). The effect of periodontal dressings on intact mucous membrane and on wound healing. A methodological study. Acta. Odontol. Scand., 8: 363. Hipp, L.I., (1978). The skin and industrial dermatosis. National Safety News, April. Jacaruso, R.B., Barlett, M.A., Carson, S. and Trombetta, L.D. (1985). Release of histamine from rat peritoneal cells in vitro as an index or irritational potential. J. Toxicol. Cut. Oculat. Toxicol., 4: 39–48. Jumblatt, M.M. and Neufeld, A.H. (1985). A tissue culture model of human corneal epithelium. In: A.M. Goldberg (Ed.), In Vitro Toxicology. Alternative Methods in Toxicology, Vol. 3. Mary Ann Liebert, New York, pp. 391–404. Kaminsky, M. and Willigan, D.A. (1982). PH and the potential irritancy of douche formulations to the vaginal mucosa of the albino rabbit and rat. Fd. Chem. Toxicol., 20: 193. Kato, I., Harihara, A. and Mizushima, Y. (1992). An in vitro method for assessing muscle irritation of antibiotics using rat primary cultured skeletal muscle fibers. Tixol. Appl. Pharmacol., 117: 194–199. Kemp, R.V., Meredith, R.W.J., Gamble, S. and Frost, M. (1983). A rapid cell culture technique for assaying the toxicity of detergent based products in vitro as a possible screen for high irritants in vivo. Cytobiol., 36: 153–159. Kemp, R.V., Meredith, R.W.J. and Gamble, S. (1985). Toxicity of commercial products on cells in suspension: A possible screen for the Draize eye irritation test. Food Chem. Toxicol., 23: 267–270. Kennah, H.E., Albulescu, D., Hignet, S. and Barrow, C.S. (1989). A critical evaluation of predicting ocular irritancy potential from an in vitro cytotoxicity assay. Fund. Appl. Toxicol., 12: 281–290.

Local Tissue Tolerance

97

Lamont, G.S., Bagley, D.M., Kong, B.M. and DeSalva, S.J. (1989). Developing an alternative to the Draize skin test: comparison of human skin cell responses to irritants in vitro. In: A.M. Goldberg (Ed.), In Vitro Toxicology: New Directions. Mary Ann Liebert, New York, pp. 183–184. Laska, D.A., Williams, P.D., Reboulet, J.T. and Morris, R.M. (1991). The L6 muscle cell line as a tool to evaluate parental products for irritation. J. Parenter. Sci. Technol., 45(2): 77–82. Lei, H., Carroll, K., Au, L. and Krag, S.S. (1986). An in vitro screen for potential inflammatory agents using cultured fibroblasts. In: A.M. Goldberg (Ed.), In Vitro Toxicology. Alternative Methods in Toxicology, Vol. 3. Mary Ann Liebert, New York, pp. 74–85. Leighton, J., Nassauer, J., Tchao, R. and Verdone, J. (1983). Development of a procedure using the chick egg as an alternative to the Draize test. In: A.M. Goldberg (Ed.), Product Safety Evaluation. Alternative Methods in Toxicology, Vol. 1. Mary Ann Liebert, New York, pp. 165–177. Lilly, G.E., Cutcher, J.L. and Henderson, M.D. (1972). Reaction of oral and mucous membranes to selected dental materials. J. Biomed. Mater. Res., 6: 545. Lindhe, J., Heyden, G., Svanberg, G., Loe, H. and Schiott, C.R. (1970). Effect of local applications of chlorhexidine on the oral mucosa of the hamster. J. Periodontal Res., 5: 177. Luepke, N.P. (1985). Hen’s egg chorioallantoic membrane test for irritation potential. Food Chem. Toxicol., 23: 287–291. Maltoltsy, A.C., Downes, A.M. and Sweeny, T.M. (1968) Studies of the investigation of the chemical nature of the water barrier, J. Invest. Dermatol., 50: 19–34. Mathias, C.G.T. (1983). Clinical and experimental aspects of cutaneous irritation. In: F. M. Margulli and H.T. Maibach (Eds.), Dermatoxicology. Hemisphere Publishing, New York, pp. 167–183. Maurice, D. and Singh, T. (1986). A permeability test for acute corneal toxicity. Toxicol. Lett., 31: 125–130. Meltzer, H.Y., Morozak, S. and Bozer, M. (1970). Effect of intramuscular injections on serum creatinine phosphokinase activty. Am. J. Med. Sci., 259: 42–48. Meyer, D.R. and McCulley, J.P. (1988). Acute and protracted injury to cornea epithelium as an indication of the biocompatability of various pharmaceutical vehicles. In: A.M. Goldberg (Ed.), Progress in In Vitro Toxicolog. Mary Ann Liebert, New York, pp. 215–235. Monash, S. and Blank, H. (1958). Location and reformation of epithelial barrier to water vapor. Arch. Dermatol., 78: 710–714. Montagna, W. (1961). The Structure and Function of the Skin, 2nd Edn. Academic Press, New York. Muir, C.K., Flower, C. and Van Abbe, N.J. (1983). A novel approach to the search for in vitro alternatives to in vivo eye irritancy testing. Toxicol. Lett., 18: 1–5. Muller, P., Raabe, G., Horold, J. and Juretzek, U. (1988). Action of chronic peracetic acid (wofasteril) administration on the rabbit oral mucosa, vaginal mucosa and skin. Exp. Pathology, 34: 223. Naughton, G.K. (1989). A physiological skin model for in vitro toxicity studies. In: Goldberg A.M. (Ed.), Alternative Methods in Toxicology, vol 7. Mary Ann Liebert, New York. Nelson, A.A., Price, C.W. and Welch, H. (1949). Muscle irritation following the injection of various penicillin preparations in rabbits. J. Am. Pharm. Assoc., 38: 237–239. Nixon, G.A., Buehler, E.V. and Newman, E.A. (1972). Preliminary safety assessment of disodium etidronate as an additive to experimental oral hygiene products. Toxicol. Appl. Pharmacol., 22: 661. North-Root, H., Yackovich, J., Demetrulias, F.J., Gucula, N. and Heinze, J.E. (1982). Evaluation of an in vitro cell toxicity test using rabbit corneal cells to predict the eye irritation potential of surfactants. Toxicol. Lett., 14: 207–212. Oliver, G.J.A. and Pemberton, M.A. (1985). An in vitro epidermal slice technique for identifying chemicals with potential for severe cutaneous effects. Food Chem. Toxicol., 23: 229–232. Prottey, C. and Ferguson, T.F.M. (1976). The effect of surfactants upon rat peritoneal mast cells in vitro. Food Chem. Toxicol., 14: 425. Reinhardt, C.A., Aeschbacher, M., Bracker, M. and Spengler, J. (1987). Validation of three cell toxicity tests and the hen’s egg test with guinea pig eye and human skin irritation data.

98

Safety Evaluation of Medical Devices

In: A.M. Goldberg (Ed.), In Vitro Toxicology—Approaches to Validation. Alternative Methods in Toxicology, Vol. 5. Mary Ann Liebert, New York, pp. 463–470. Reinhardt, C.A., Pelli, D.A. and Zbinden, G. (1985). Interpretation of cell toxicity data for the estimation of potential irriation. Food Chem. Toxicol., 23: 247–252. Scaife, M.C. (1982). An investigation of detergent action on in vitro an possible correlations with in vivo data. Int. J. Cosmet. Sci., 4: 179–193. Selling, J. and Ekwall, B. (1985). Screening for eye irritancy using cultured HeLa cells. Xenobiotica, 15: 713–717. Shadduck, J.A., Everitt, J. and Bay, P. (1985). Use of in vitro cytotoxicity to rank ocular irritation of six surfactants. In: A.M. Goldberg (Ed.), In Vitro Toxicology. Alternative Methods in Toxicology. Vol. 3. Mary Ann Liebert, New York, pp. 641–649. Shadduck, J.A., Render, J., Everitt, J., Meccoli, R.A. and Essexsorlie, D. (1987). An approach to validation: comparison of six materials in three tests. In: A.M. Goldberg (Ed.), In Vitro Toxicology— Approaches to Validation. Alternative Methods in Toxicology, Vol. 5. Mary Ann Liebert, New York, pp. 75–78. Shintani, S., Yamazaki,, M., Nakamura, M. and Nakayama, I. (1967). A new method to determine the irritation of drugs after intramuscular injections in rabbits. Tox. Appl. Pharm., 11: 293–301. Shopsis, C. and Sathe, S. (1984). Uridine uptake inhibition as a cytotoxicity test: Correlation with the Draize test. Toxicology, 29: 195–206. Sidell, F.R., Claver, D.L. and Kaminskis, A. (1974). Serum creatine phosphokinase activity after intramuscular injection. JAMA, 228: 1884–1887. Simons, P.J. (1981). An alternative to the Draize test. In: A.N. Rowan and C.J. Startmann (Eds.), The use of Alternatives in Drug Research. MacMillan Press, London. SOT Position Paper (1989). Comments on the LD 50 and acute eye and skin irritation tests. Fundam. Appl. Toxicol., 13: 621–623. Soto, R.J., Servi, M.J. and Gordon, V.C. (1988). Evaluation of an alternative method of ocular irritation. In: A.M. Goldberg (Ed.),

Progress in In Vitro Toxicology. Mary Ann Liebert, New York, pp. 289–296. Swisher, D.A., Prevo, M.E. and Ledger, P.W. (1988). The MTT in vitro cytotoxicity test: correlation with cutaneous irritancy in two animal models. In: A.M. Goldberg (Ed.), Progress in In Vitro Toxicology. Mary Ann Liebert, New York, pp. 265–269. Tchao, R. (1988). Trans-epithelial permeability of fluorescein in vitro as an assay to determine eye irritants. In: Goldberg A.M. (Ed.), Alternative Methods in Toxicology, Vol. 6. Mary Ann Liebert, New York, pp. 271–283. USP (2007). Pyrogen test. United States Pharmacopeia, XXX. USP Convention, Rockville, Maryland, p. 1515. USP (2007). Intramuscular irritation test. United States Pharmacopeia, XXX. USP Convention, Rockville, Maryland, pp. 1180–1183. USP (2007). Bacterial endotoxins test. United States Pharmacopeia, XXX. USP Convention, Rockville, Maryland, pp. 1493–1495. Watanabe, M., Watanabe, K., Suzuki, K. et al. (1988). In vitro cytotoxicity test using primary cells derived from rabbit eye is useful as an alternative for Draize testing. In: A.M. Goldberg (Ed.), Progress in In Vitro Toxicology. Mary Ann Liebert, New York, pp. 285–290. Weary, M. and Baker, B. (1977). Utilization of the limulus amebocyte lysate test for pyrogen testing of large volume parenterals, administration sets and medical devices. Bull. Parenter. Drug Assoc., 31: 1127–133. Weil, C.S. and Scala, R.A. (1971). Study of intra-and interlaboratory variability in the results of rabbit eye and skin irritation tests. Toxicol. Appl. Pharmacol., 19: 276–360. Williams, P.D., Masters, B.G., Evans, L.D., Laska, D.A. and Hattendorf, G.H. (1987). An in vitro model for assessing muscle irritation due to parenteral antibiotics. Fundam. Appl. Toxicol. 9: 10–17. Young, M.F., Trobetta, L.D. and Sophia, J.V. (1986). Correlative in vitro and in vivo study of skeletal muscle irritancy. Toxicologist, 6(1): 1225.

9 Immunotoxicology

The evaluation of the immunotoxicity of medical devices as part of their biocompatibility assessment is the subject of the most recent ISO-10993 guidance (part 20). Traditionally, and still the case for most skin only contact devices, delayed contact, simple dermal sensitization was adequate for evaluating this endpoint. This is in spite of the fact that the association between implanted or indwelling devices and granuloma formation has been known for some time (Adams, 1953; Anderson, 1988; Black, 1981; Burkett et al., 1986; Woodward and Salthouse, 1986; Unanue, 1994; Salthouse, 1982; Marchant et al., 1985). However, improved science and the record of immune based device problems since 1980 has brought the adequacy of this approach into question. The immune system is a highly complex system of cell tissues and mediators involved in a multitude of functions, including antigen presentation and recognition, amplification, and cell proliferation with subsequent differentiation and secretion of lymphokines and antibodies (Bick, 1985). The end result is an integrated and highly interdependent system responsible for defense against foreign pathogens and spontaneously occurring neoplasms that, if left unchecked, may result in infection and malignancy. We also now speak of the system being composed of innate (present in both invertebrates and vertebrates) and adaptive (present only in vertebrates) components. To be effective, the immune system must be able to both recognize and destroy foreign antigens. To accomplish this, cellular and soluble components of the adaptive immune system of diverse function and specificity circulate through blood and lymphatic vessels, thus allowing them to act at remote sites and tissues. For this system to function properly requires regulation through cell-to-cell communications and precise recognition of self versus nonself (or threat/nonthreat). Immunotoxicants can upset this balance if they are lethal to one of more of the cell types or alter membrane morphology and receptors. There are several undesired immune system responses that may potentially occur upon repeated exposure to a medical device material that may ultimately present barriers to its development, including: ●

●

●

●

Immunotoxicology has evolved since the late 1970s as a specialty within toxicology that brings together knowledge from basic immunology, molecular biology, microbiology, pharmacology, and physiology. As a discipline, immunotoxicology involves the study of the adverse effects that xenobiotics have on the immune system. As listed above, several different types of adverse immunological effects may occur, including immunosuppression, autoimmunity, and hypersensitivity. Although these effects are clearly distinct, they are not mutually exclusive. For example, immunosuppressive drugs that suppress suppressor-cell activity can also induce autoimmunity (Hutchings et al., 1985), and agents that are immuno-enhancing at low doses may be immunotoxic at high doses. Chemical xenobiotics may be in the form of natural or man-made environmental chemicals— pharmaceuticals and biologicals that are pharmacologically, endocrinologically, or toxicologically active. Although, in general, xenobiotics are not endogenously produced, immunologically active biological response modifiers that naturally occur in the body should also be included, since many are now known to compromise immune function when present in pharmacologically effective doses (Koller, 1987). The success in development of immune modulary protein therapeutics likewise has had to both promise and challenge. Although the types of immunological responses to various xenobiotics may be similar, the approach taken for screening potential immunological activity will vary depending on the application of the compound. In contrast to potential environmental exposures, medical devices and the materials they are composed of are developed with intentional but restricted human exposure and their biological effects are extensively studied in surveillance. In 1993, the FDA issued draft guidelines for immunotoxicity testing in the revision of the “Redbook” (FDA, 1993). Although these guidelines have been established through the Center for Food Safety and Applied Nutrition, other centers within the FDA extended the usage of these guidelines to cover testing for human and veterinary pharmaceuticals. Both ICH (for pharmaceuticals) and ISO (for drugs) have promulgated guidances for the evaluation of subject products for adverse immune effects. Unanticipated immunotoxicity is infrequently observed with drugs that have been approved for marketing. With the exception of drugs that are intended to be immunomodulatory or immunosuppressive as part of their therapeutic mode of action, there is little evidence that drugs or devices cause unintended functional immunosuppression in man (Gleichman et al., 1989). However, hypersensitivity (allergy) and autoimmunity are frequently observed and are serious consequences of some therapies (DeSwarte, 1986). An adverse immune response in the form of hypersensitivity is one of the most frequent safety causes for withdrawal of drugs that have already made it to market (see Table 9.1) and accounts for approximately 15% of adverse reactions to

Down-modulation of the immune response (immunosuppression or hypo immunity), which may result in an impaired ability to deal with neoplasia and infections. This is of particular concern if the device is intended or likely to be used in patients with pre-existing conditions such as cancer, severe infection, or immunodeficiency diseases. Up-modulation of the immune system (or hyper immunity, such as auto-immunity). Direct adverse immune responses to the agent itself in the form of hypersensitivity responses (anaphylaxis and delayed contact hypersensitivity). Direct immune responses to the device that limit or nullify its utility (i.e., the development of neutralizing antibodies to be a delivered agent). 99

100

Safety Evaluation of Medical Devices

Table 9.1 Drugs Withdrawn from the Market Due to Dose- and Time-Unrelated Toxicity not Identified in Animal Experiments Compound

Adverse reaction

Aminopyrine Phenacetin Dipyrone Clioquinol

Agranulocytosis Intersitial nephritis Agranulocytosis Subacute myelo-optic neuropathy Oxyphenisatin Chronic active hepatitis Nialamide Liver damage Phenoxyorioazine Liver damage Mebanazine Liver damage Ibufenac Hepatotoxicity Practolol Oculo-mucocutaneous syndrome Alclofenace Hypersensitivity Azaribine Thrombosis Ticynafen Nephropathy Benoxaprofen Photosensitivity, hematoxicity Zomepirac Urticaria, anaphylactic shock Zirnelidine Hepatotoxicity

Year of introduction

Years on the market

Approx 1900 Approx 1900 Approx 1930 Approx 1930

75 83 47 51

Approx 1955 1959 1961 1963 1966 1970

23 19 5 3 2 6

1972 1975 1979 1980

7 1 1 2

1980

3

1982

2

OVERVIEW OF THE IMMUNE SYSTEM

Source: Adapted from Bakke et al. (1984).

xenobiotics (de Weck, 1983). In addition, adverse immune responses such as this (usually urticaria and frank rashes) are the chief “unexpected” finding in clinical studies. These findings are unexpected in that they are not predicted by preclinical studies. These findings are unexpected in that they are not predicted by preclinical studies because there is a lack of good preclinical models for predicting systemic hypersensitivity responses, especially to orally administered agents. Table 9.2

As a consequence, the unexpected occurrence of hypersensitivity in the clinic may delay, or even preclude, further development and commercialization. Thus, a primary purpose for preclinical immunotoxicology testing is to help us detect these adverse effects earlier in development, before they are found in clinical trials.

A thorough review of the immune system is not the intent of this chapter, but a brief description of the important components of the system and their interactions is necessary for an understanding of how xenobiotics can affect immune function. A breakdown at any point in this intricate and dynamic system can lead to immunopathology. The immune system is divided into two defense mechanisms: nonspecific, or innate, and specific, or adaptive, mechanisms that recognize and respond to foreign substances. Some of the important cellular components of nonspecific and specific immunity are described in Table 9.2. The nonspecific immune components are the phagocytic cells such as the monocytes, macrophages, and polymorphic neutrophils (PMNs). The specific, or adaptive, immune system is characterized by memory, specificity, and the ability to distinguish “self” from “nonself” (Battisto et al., 1983), though an alternative to this basic self–nonself paradigm has been proposed that rater the immune system actually responds to some form of “danger” manager (Pennisi, 1996). The important cells of the adaptive immune system are the lymphocytes and antigen-presenting cells that are part of nonspecific immunity. The lymphocytes, which originate from pluripotent stem cells located in the hematopoietic tissues of the liver (fetal) and bone marrow, are composed of two general cell types;

Cellular Components of the Immune System and their Functions

Cell subpopulations Nonspecific immunity Granulocytes Neutrophils (blood) Basophils (blood) Eosinophils (blood) Mast cells (connective tissue) Natural killer cells (NK) Retucykiendothelial

Markersa

Functions Degranulate to release mediators

CD14; HLA-DR

Nonsensitized lymphocytes; directly kill target cells Antigen processing, presentation, and phagocytosis (humoral and some cell-mediated responses)

Macrophage (peritoneal, pleural, alveolar spaces) Histiocytes (tissues) Monocytes (blood) Specific immunity Humoral immunity Activated B cells Plasma cells Resting Memory Cell-mediated immunity T-Cell types: Helper (Tk) Cytotoxic (Tk) Suppressor (Ts) a

CD19; CD23

Proliferate; form plasma cells Secrete antibody; terminally differentiated Secrete IgM antibodies (primary response) Secrete IgG antibodies (secondary response)

CD4; CD25 CD8; CD25 CD8; CD25

Assists in humoral immunity; required for antibody production Targets lysis Suppresses/regulates humoral and cell-medicated responses

Activation surface markers detected by specific monoclonal antibodies; can be assayed with flow cytometry.

Chapter 9:

Table 9.3 Ag (1)

Antibodies

Serum conc. mg mL−1 (%)

IgG

10–12 (80%)

IgM

1–2 (5–10%)

IgA

3–4 (10–15%)

IgD

0.03 (< 1%)

IgE

< 0.0001

(2) T cell

IL-1

IL-2 Tk

(3)

(5)

B cell Ab

Th

Ag (4)

101

Antibodies Involved in the Humoral Immune Response

APC (macrophage)

Immunotoxicology

Characteristics/functions Monomeric structure (γ-globulin); secreted from B cells during secondary response; binds complement; can cross placenta Pentameric structure; secreted from B cells during primary response; potent binder of complement; high levels indicative of systemic lupus erythematosus or rheumatoid arthritis, cannot across placenta. Dimeric or monomeric structures; found in seromucous secretions (breast milk); secreted by B cells associated with epithelial cells in GI tract, lung, etc. Monomer; extremely labile; functions not well known. Reaginic antibody involved in immediate hypersensitivity; antihelminthic; does not bind complement.

Source: Extracted and modified from Clark (1983). Ts ILs

Figure 9.1 A simplified schematic of the immunoregulatory circuit that regulates the activation of T cells and B cells involved in humoral (T-cell dependent) and cell-mediated immunity. (1) Antigen (Ag) is processed by the APCs expressing class II MHC molecules. (2) Antigen plus class II MHC is then presented to antigen-specific T helper cells (CD4+) which stimulate secretion of IL02. (3) IL-2 in turn stimulates proliferation (clonal expansion) of T cells and differentiation into T suppressor (Ts), T killer (Tk), and T helper (Th) effector cells. The expanded clone has a higher likelihood of finding the appropriate B cell that has the same antigen and class II molecules on its surface. (4) Next, the antigen binds to an antibody (Ab) on the surface of a specific B cell. (5) The B cell, in turn, processes the antigen and presents it (plus class II MHC) to the specific Th cell. The Th cell is then stimulated to secrete additional interleukins (ILs) that stimulate clonal expansion and differentiation of the antigen-specific B cell.

T and B cells. The T cells differentiate in the thymus and are made up of three subsets: helper, suppressor, and cytotoxic. The B cells, which have the capacity to produce antibodies, differentiate in the bone marrow or fetal liver. The various functions of the T cells include presenting antigen to B cells, helping B cells to make antibody, killing infected cells, regulating the level of the immune response, and stimulating cytotoxic activity of other cells such as macrophages (Male et al., 1987). Activation of the immune system is thought to occur when antigen-presenting cells (APCs) such as macrophages and dendritic cells take up antigen via Fc or complement receptors, process the antigen, and present it to T cells (see Figure 9.1). Macrophages release soluble mediators such as interleukin 1 (IL-1), which stimulate T cells to proliferate. Antigen-presenting cells must present antigen to T cells in conjunction with the class II major histocompatibility complex (MHC) proteins that are located on the surface of the T cells. The receptor proteins, and the T3 cell is a complex of the Ti molecule that binds antigen, the MHC proteins, and

the T3 molecular complex, which is often referred to as the CD3 complex. Upon stimulation, T cells proliferate, differentiate, and express interleukin-2 (IL-2) receptors. T cells also produce and secrete IL-2, which, in turn, acts on antigenspecific B cells, causing them to proliferate and differentiate into antibody-forming (plasma) cells. Antibodies (Table 9.3) circulate freely in the blood or lymph and are important in the neutralizing foreign antigens. The various types of antibodies involved in humorgenes (polymorphisms) that encode diversity to the variable region of the antibody. B cells are capable of generating further diversity to antibody specificity by a sequence of molecular events involving somatic mutations, chromosomal rearrangements during mitosis, and recombinations of gene segments (Roitt et al., 1985). The immune system is regulated in part by feedback inhibition involving complex interactions between the various growth and differentiation factors listed in Table 9.4. Since antigen initiates the signal for the immune response, elimination of antigen will decrease further stimulation (Male et al., 1987). T suppressor cells (Ts) also regulate the

Table 9.4 Cell type Tk cells

Cells and Mechanisms Involved in Cell-Mediated Cytotoxicity Mechanism of cytotoxicity

Tk cells that are specifically sensitized to antigens on target cells interact directly with target cells to lyse them. Cells involved in delayed hypersensitivity that act indirectly TD cells to kill target cells; TD cells react with antigen and release cytokines that can kill target cells. NK cells Nonspecific T cells that react directly with target cells (tumor cells) without prior sensitization. Null cells Antibody-dependent cell-mediated cytotoxicity (ADCC) involving non-T/non-B cells (null cells) with Fc receptors specific fir antibody-coated target cells. Macrophages Nonspecific, direct killing of target by phagocytosis; also involved in presenting antigen to specific Tk cells that can then mediate cytotoxicity as described above.

102

Safety Evaluation of Medical Devices

Table 9.5

Growth and Differentiation Factors of the Immune System

Factors Interleukins IL-1 IL-2

Cell of origin

Primary immune functions

Macrophage, B and T cells T cells (Th)

Lymphocyte-activating factor; enhances activation of T and B cells, NK cells, and macrophages T-cell growth factor; stimulates T-cell growth and effector differentiation; stimulates B-cell proliferation/differentiation Mast-cell growth factor; stimulates proliferation/differentiation of mast cells, neutrophils, and macrophages B-cell growth factor; induces proliferation/differentiation of B cells and secretion of IgA, IgG1, and IgE; promotes T-cell growth; activates macrophages Stimulates growth/differentiation of B cells and secretion of IgG; promotes IL-2-induced growth of T cells Stimulates pre-B- and pre-T-cell growth/differentiation; enhances thymocyte proliferation Neutrophils chemotaxis Stimulates T cells and mast cells Stimulates mast cells and thymocytes; induction of class II MHC

a

IL-3

T cells (Th)

IL-4

T cells (Th), mast cells, B cells

IL-6

T cells, fibroblasts, monocytes

IL-7 IL-8 IL-9 IL-10

Bone marrow stromal cells Monocytes, fibroblasts T cells T cells

Interferons (INF) α-INF β-INF γ-INF

Leukocytes and mast cells Fibroblasts, epithelial cells T cells (Th), cytotoxic T cells

Tumor necrosis factors (TNF) TNFα Macrophage, B and T cells TNFβ

T cells (Th) NK cells

Colony stimulating factors (CSF) Stem cells: Granulocyte CSF Myeloid Macrophage CSF Myeloid GranulocytesMyeloid macrophage CSF

Antiviral; increases NK-cell function, B-cell differentiation, potentiates macrophage production of IL-1 Antiviral; potentiates macrophage production of IL-1; increases NK-cell function Antiviral; activates macrophages; induces MHC class II expression on macrophages, epithelial, and endothelial cells Catectin; promotes tumor cytotoxicity; activates macrophages and neutrophils; enhances IL-2 receptor expression on T cells; inhibits antibody secretion Lymphotoxin; promotes T-cell-mediated cytotoxicity B cell activation Promotes growth and differentiation of: Granulocytes and macrophages Macrophages and granulocytes Granulocytes, macrophages, eosinophils, mast cells, and pluripotent progenitor cells

a Includes lymphokines, monokines, and cytokines produced by T cells, macrophages, and other cells, respectively. Source: Extracted and modified from Golub and Green (1991).

immune response and are thought to be important in the development of tolerance to self antigens. In addition to the humoral immune system or the branch that is modulated by antibody, cell-mediated immunity and cytotoxic cell types play a major role in the defense against virally infected cells, tumor cells, and cells of foreign tissue transplants. Cytotoxic Tk cells (T killer cells) recognize antigen in association with class I molecules in MHC, while natural killer cells (NK cells) are not MHC effector cell, lysosomal degranulation, calcium influx into the targeted cell. The various types of cells involved in cell-mediated cytotoxicity and their mechanisms of action are outlined in Table 9.5.

IMMUNOTOXIC EFFECTS The immune system is a highly integrated and regulated network of cell types that requires continual renewal to achieve balance and immunocompetence. Fortunately, the multiple components of the system serve to overlap and respond to challenges in a manner which provides redundancy. The delicacy of this balance makes specific components of the immune system a natural target for cytotoxic drugs or their metabolites. Since renewal is dependent on the ability of cells to proliferate and differentiate, exposure to agents that arrest cell division can subsequently lead to reduced immune function or immonosuppression. This concept

has been exploited in the development of therapeutic drugs intended to treat leukemias, auto-immune disease, and chronic inflammatory diseases and to prevent transplant rejection. However, some drugs adversely modulate (or overly stimulate) the immune system secondarily to their therapeutic effects. (Simply consider the case of the monoclonal antibody TGN.) Two broad categories of immunotoxicity have been defined on the basis of suppression or stimulation of normal immune function. Immunosupression is a down-modulation of the immune system characterized by cell depletion, dysfunction, or dysregulation that may subsequently result in increased susceptibility to infection and tumors. By contrast, immunostimulation is an increased or exaggerated immune responsiveness that may be apparent in the form of a tissuedamaging allergic hypersensitivity response or pathological autoimmunity. However, as knowledge of the mechanisms involved in each of these conditions has expanded, the distinction between them has become less clear. Some agents can cause immunosuppression at one dose or duration of exposure, and immunostimulation at others. For instance, the chemotherapeutic drug cyclophosphamide is in most cases immunosuppressive; however, it can also induce autoimmunity (Hutchings et al., 1985). Likewise, dimethylnitrosamine, a nitrosamine detected in some foods, has been shown to have both suppressing and enhancing effects on the immune system (Yoshida et al., 1989).

Chapter 9:

IMMUNOSUPPRESSION The various cells of the immune system may differ in their sensitivity to a given xenobiotic. Thus, immunosuppression may be expressed as varying degrees of reduced activity of a single cell type or multiple populations of immunocytes. Several lymphoid organs such as the bone marrow, spleen, thymus, and lymph nodes may be affected simultaneously or the immunodeficiency may be isolated to a single tissue, such as the Peyer’s patches of the intestines. The resulting deficiency may in turn lead to an array of clinical outcomes of varying ranges of severity. These outcomes include increased susceptibility to infections, increased severity or persistence of infections, or infections with unusual organisms (e.g., system fungal infections). Immunosuppression can be induced in a dose-related manner by a variety of therapeutic agents at dose levels lower than those required to produce overt clinical signs of general toxicity. In addition, immunosuppression can occur without regard to genetic predisposition, given that a sufficient dose level and duration of exposure has been achieved. Humoral immunity is characterized by the production of antigen-specific antibodies that enhance phagocytosis and destruction of microorganisms through opsonization. Thus, deficiencies of humoral immunity (B lymphocytes) may lead to reduced antibody titers and are typically associated with acute gram positive bacterial infections (i.e., Streptococcus). Although chronic infection is usually associated with dysfunction of some aspect of cellular immunity, chronic infections can also occur when facultative intracellular organisms such as Listeria or Mycobacterium evade antibodies and multiply within phagocytic cells. Since cellular immunity results in the release of chemotactic lymphocytes that in turn enhance phagocytosis, a deficiency in cellular immunity may also result in chronic infections. Cellular immunity is mediated by T cells, macrophages, and NK cells involved in complex compensatory networks and secondary changes. Immunosuppressive agents may act directly by cytotoxicity to T cells, or indirectly by blocking mitosis, lymphokine synthesis, lymphokine release, or membrane receptors to lymphokines. In addition, cellular immunity is involved in the production and release of interferon, a lymphokine that ultimately results in blockage of viral replication (Table 9.5). Viruses are particularly susceptible to cytolysis by T cells since they often attach to the surface of infected cells. Thus, immunosuppression of any of the components of cellular immunity may result in an increase in protozoan, fungal, and viral infections as well as opportunistic bacterial infections. Immune depression may result unintentionally as a side effect of cancer chemotherapy or intentionally from therapeutics administered to prevent graft rejection. In fact, both transplant patients administered immunosuppressive drugs and cancer patients treated with chemotherapeutic agents have been shown to be at high risk of developing secondary cancers, particularly of lymphoreticular etiology (Penn, 1977). Most of these drugs are alkylating or crosslinking agents that by their chemical nature are electrophilic and highly reactive with nucleophilic macromolecules (protein and nucleic acids). Nucleophilic sites are quite ubiquitous and include amino, hydroxyl, mercapto, and histidine functional groups. Thus, immunotoxic agents used in chemotherapy may induce secondary tumors through direct genotoxic mechanisms (i.e., DNA alkylation). Reduced cellular immunity may result in increased malignancy and decreased viral resistance through indirect

Immunotoxicology

103

mechanisms as well, by modulating immune surveillance of aberrant cells. T lymphocytes, macrophage cells, and NK cells are all involved in immunosurveillance through cytolysis of virally inflected cells or tumor cells, each by a different mechanism (Table 9.2) (Burnet, 1970). In addition to the common cell types described in Table 9.2, at least two other types of cytotoxic effector cells of T-cell origin have been identified, each of which has a unique lytic specificity phenotype, and activity profile (Merluzzi, 1985). Of these, both LAK and TIL cells have been shown to lyse a variety of different tumor cells. However, TIL cells have 50–100 times more lytic activity than LAK cells. Most tumor cells express unique surface antigens that render them different from normal cells. Once detected as foreign, they are presented to the T helper cells in association with MHC molecules to form an antigen–MHC complex. This association elicits a genetic component to the immunospecificity reaction. T helper cells subsequently direct the antigen complex toward the cytotoxic T lymphocytes, which possess receptors for antigen–MHC complexes. These cells can then proliferate, respond to specific viral antigens or antigens on the membranes of tumor cells, and destroy them (Yoshida et al., 1989). In contrast, the macrophages and natural killer (NK) cells are involved in nonspecific immunosurveillance in that they do not require prior sensitization with a foreign antigen as a prerequisite for lysis, and are not involved with MHC molecules. The enhancement of either NK cell function or macrophage function has been shown to reduce metastasis of some types of tumors. Macrophage cells accumulate at the tumor site and have been shown to lyse a variety of transformed tumor cells (Volkman, 1984). Natural killer cells are involved in the lysis of primary autochthonous tumor cells. Migration of NK cells to tumor sites has been well documented. Although not clearly defined, it appears that they can recognize certain proteinaceous structures on tumor cells and lyse them with cytolysin.

IMMUNOSTIMULATION A variety of drugs as well as environmental chemicals have been shown to have immunostimulatory or sensitizing effects on the immune system and these effects are well documented in humans exposed to drugs (DeSwarte, 1986). The drug or metabolite can act as a hapten and covalently bind to a protein or other cellular constituent of the host to appear foreign and become antigenic. Haptens are low molecular weight substances that are not in themselves immunogenic but will induce an immune response if conjugated with nuclophilic groups on proteins or other macromolecular carriers. In both allergy and autoimmunity, the immune system is stimulated or sensitized by the drug conjugate to produce specific pathological responses. An allergic hypersensitivity reaction may vary from one which results in an immediate anaphylactic response to one which produces a delayed hypersensitivity reaction or immune complex reaction. Allergic hypersensitivity reactions result in a heightened sensitivity to nonself antigens, whereas autoimmunity results in an altered response to self antigens. Unlike immunosuppression, which nonspecifically affects all individuals in a dose-related manner, both allergy and autoimmunity have a genetic component that creates susceptibility in those individuals with a genetic predisposition. Susceptible individuals, once sensitized, can respond to genetic predisposition and can respond to even minute quantities of the antigen.

104

Safety Evaluation of Medical Devices

Table 9.6

Types of Hypersensitivity Responses

Type and designationa

Components

Effects

Mechanism

I. Immediate

Mast cells; IgE

II. Cytoxic

Antibodies

Anaphylaxis, asthma, urticar uriticaria, rhinitis, dermatitis Hemolytic anemia, Goodpasture’s disease

III. Immune complex (arthus) IV. Delayed hypersensitivity

Antigen–antibody complexes (Ag–Ab) TD cells; macrophages

IgE binds to mast cells to stimulate release of humoral factors IgG and IgM bind to cells (e.g., RBC’s), fix complement (opsinization), then lyse cells Ag–Ab complexes deposit in tissues, and may fix complement Sensitized T cells induce a delayed hypersensitivity response upon challenge

SLE rheumatoid arthritis, glomerular nephritis, serum sickness, vasculitis Contact dermatitis, tuberculosis

Source: Based on classification system of Gell and Coombs (1975).

HYPERSENSITIVITY The four types of hypersensitivity reactions (“sensitization”), as classified by Coombs and Gell (1975), are outlined in Table 9.6. The first three types are immediate antibody-mediated reactions, whereas the fourth type is a cellular-mediated delayed-type response that may require 1–2 days to occur after a secondary exposure. Type I reactions are characterized by an anaphylaxis response to a variety of compounds, including proteinaceous materials and pharmaceuticals such as penicillin. Various target organs may be involved depending on the route of exposure. For example, the gastrointestinal tract is usually involved with food allergies, the respiratory system with inhaled allergens, the skin with dermal exposure, and smooth muscle vasculature with systemic exposure. The type of response elicited often depends on the site of exposure and includes dermatitis and urticaria (dermal), rhinitis and asthma (inhalation), increased gastrointestinal emptying (ingestion), and systemic anaphylactic shock (parenteral). Type I Hypersensitivity. During an initial exposure, IgE antibodies are produced and bind to the cell surface of mast cells and basophils. Upon subsequent exposures to the antigen, reaginic IgE antibodies bound to the surface of target cells at the Fc region (mast cells and basophils) become cross-linked (at the Fab regions) by the antigen. Cross-linking causes distortion of the cell surface and IgE molecule, which, in turn, activates a series of enzymatic reactions, ultimately leading to degranulation of the mast cells and basophils. These granules contain a variety of pharmacological substances (Table 9.7), such as histamines, serotonins, prostaglandins, bradykinins, and leukotrienes (SRS-A and ECR-A). Upon subsequent challenge exposures, these factors are responsible for eliciting an allergic reaction through vasodilation and increased vascular permeability. The nasal passages contain both mast cells and plasma cells that secrete IgE antibodies. Allergic responses localized in the nasal mucosa result in dilation of the local blood vessels, tissue swelling, mucus secretion, and sneezing. Reactions localized in the respiratory tract, also rich in mast cells and IgE, result in allergic asthma response. This condition is triggered by the release of histamine and SRS-A, which induce constriction of the Bronchi and alveoli, pulmonary edema, and mucous secretions that block the bronchi and alveoli, together resulting in severe difficulty in breathing. In the case of a challenge dose of a drug administered systemically, the reactive patient may have difficulty breathing within minutes of exposure and may experience convulsions, vomiting, and low blood pressure. The effects of anaphylactic shock and respiratory distress, if severe, may ultimately result in death.

Antibiotics containing β-lactam structures, such as penicillin and cephalosporins, are the most commonly occurring inducers of anaphylactic shock and drug hypersensitivity in general. Other hypersensitivity reactions may include urticarial rash, fever, bronchospasm, serum sickness, and vasculitis with reported incidences of all types varying from 0.7 to 10% (Idsøe et al., 1968) and the incidence of anaphylactoid reactions varying from 0.04 to 0.2%. When the β-lactam ring is opened during metabolism, the penicilloyl moiety can form covalent conjugates with nucleophilic sites on proteins. The penicilloyl conjugates can then act as haptens to form the determinants for antibody induction. Although most patients that have received penicillin produce antibodies against the metabolite benzylpenicilloyl, only a fraction experience allergic reaction (Garratlz and Petz, 1975), which suggests a genetic component to susceptibility. Type II Hypersensitivity. Type II cytolytic reactions are mediated by IgG and IgM antibodies that can fix complement, opsonize particles, or induce antibody-dependent cellular cytolysis reactions. Erythrocytes, lymphocytes, and platelets of the circulatory system are the major target cells that interact with the cytolytic antibodies, causing depletion of these cells. Hemolytic anemia (penicillin, methyldopa), leukopenia, thrombocytopenia (quinidine), and/or granulocytopenia (sulfonamide) may result. Type II reactions involving the lungs and kidneys occur through the development of antibodies (autoantibodies) to the basement membranes in the alveoli or glomeruli, respectively. Prolonged damage may result in Goodpasture’s disease, an autoimmune disease characterized by pulmonary hemorrhage and glomerulonephritis. Several other autoimmune-type

Table 9.7

Proteins and Soluble Mediators Involved in Hypersensitivity

Factor

Origin

Characteristics/functions

Histamine

SRS-A

Mast cells, basophils Mast cells, basophils Lung tissue

ECF-A

Mast cells

Prostaglandins

Various tissues

Contraction of smooth muscle; increases vascular permeability Contraction of smooth muscle; leukotriene (Slow-reacting substance of anaphylaxis): Contraction of smooth muscle; acidic polypeptide (Eosinophilic chemotactic factor of anaphylaxis); attracts eosinophils; small peptide Modifies release of histamine and serotonin from mast cells and basophils

Serotonin

Source: Extracted and modified from Clark (1983).

Chapter 9:

diseases have been associated with extended treatments with D-penicillamine and other pharmaceuticals. Various types of autoimmune responses and examples of drug-induced autoimmunity are discussed in further detail later in this section. Type III Hypersensitivity. Type III reactions (arthus) are characterized as an immediate hypersensitivity reaction initiated by antigen–antibody complexes that form freely in the plasma instead of at the cell surface. Regardless of whether the antigens are self or foreign, complexes mediated by IgG can form and settle into the tissue compartments of the host. These complexes can then fix complement and release C3a and C5a fragments that are chemotactic for phagocytic cells. Polymorphonuclear leukocytes are then attracted to the site; where they phagocytize the complexes and release hydrolytic enzymes into the tissues. Additional damage can be caused by binding to and activating platelets and basophils, which, in the end, results in localized necrosis, hemorrhage, and increased permeability of local blood vessels. These reactions commonly target the kidney, resulting in glomerulonephritis through the deposition of the complexes in the glomeruli. Some devices and antibiotics (β-lactam) have been reported to produce glomerular nephritis in humans that has been attributed to circulating immune complexes. These complexes have also been observed in preclinical toxicology studies with baboons treated with a β-lactam antibiotic, prior to the appearance of any biochemical or clinical changes (Descotes and Mazue, 1987). In addition, immunoglobulin complexes have been observed in rats treated with gold and autologous immune complex nephritis has been observed in guinea pigs (Ueda et al., 1980). Similar evidence of immunomediated nephrotoxicity has been reported in rheumatoid arthritis patients administered long-term treatments with gold compounds; proteinuria has been observed in approximately 10% of these patients. Other target organs such as the skin with lupus, the joints with rheumatoid arthritis, and the lungs with pneumonitis may be affected. The deposition of antigen–antibody complexes through the circulatory system results in a syndrome referred to as serum sickness, which was quite prevalent prior to 1940 (Clark, 1983), when serum for diphtheria was commonly used. Serum sickness occurs when the serum itself becomes antigenic as a side effect from passive immunization with heterologous antiserum produced from various sources of farm animals. The antitoxin for diphtheria was produced in a horse and administered to humans as multiple injections of passive antibody. As a consequence, these people often became sensitized to the horse serum and developed a severe form of arthritis and glomerulonephritis caused by deposition of antigen–antibody complexes. Clinical symptoms of serum sickness present as urticarial skin eruptions, arthralgia or arthritis, lymphadenopathy, and fever. Drugs such as sulfonamides, penicillin, and iodides can induce a similar type of reaction. Although uncommon today, transplant patients receiving immunosuppressive therapy with heterologous antilymphocyte serum or globulins may also exhibit serum sickness (Ueda et al., 1980). Type IV Delayed-type Hypersensitivity (DTH). Delayedtype hypersensitivity reactions are T-cell mediated with no involvement of antibodies. However, these reactions are controlled through accessory cells, suppressor T cells, and monokine-secreting macrophages, which regulate the proliferation and differentiation of T cells. The most frequent form

Immunotoxicology

105

of DTH manifests itself as contact dermatitis. The drug or metabolite binds to a protein in the skin or the Langerhans cell membrane (class II MHC molecules) where it is recognized as an antigen and triggers cell proliferation. After a sufficient period of time for migration of the antigen and clonal expansion (latency period), a subsequent exposure will elicit a dermatitis reaction. A 24–48 hr delay often occurs between the time of exposure and onset of symptoms to allow time for infiltration of lymphocytes to the site of exposure. The T cells (CD4+) that react with the antigen are activated and release lymphokines that are chemotactic for monocytes and macrophages. Although these cells infiltrate to the site via the circulatory vessels, an intact lymphatic drainage system from the site is necessary since the reaction is initiated in drainage lymph nodes proximal to the site (Clark, 1983). The release (degranulation) of enzymes and histamines from the macrophages may then result in tissue damage. Clinical symptoms of local dermal reactions may include a rash (not limited to sites of exposure) itching, and/or burning sensations. Erythema is generally observed in the area around the site, which may become thickened and hard to the touch. In severe cases, necrosis may appear in the center of the site followed by desquamation during the healing process. The immune-enhancing drugs isoprinosine and avridine have been shown to induce a delayed-type hypersensitivity reaction in rats (Exon et al., 1986). A second form of delayed-type hypersensitivity response is similar to that of contact dermatitis in that macrophages are the primary effector cells responsible for stimulating CD4+ T cells; however this response is not necessarily localized to the epidermis. A classical example of this type of response is demonstrated by the tuberculin diagnostic tests. To determine if an individual has been exposed to tuberculosis, a small amount of fluid from tubercle bacilli cultures is injected subcutaneously. The development of induration after 48 hr at the site of injection is diagnostic of prior exposure. Shock, similar to that of anaphylaxis, may occur as a third form of a delayed systemic hypersensitivity response. However, unlike anaphylaxis, IgE antibodies are not involved. This type of response may occur 5–8 hr after systemic exposure and can result in fatality within 24 hr following intravenous or intraperitoneal injection. A fourth form of delayed hypersensitivity results in the formation of granulomas. If the antigen is allowed to persist unchecked, macrophages and fibroblasts are recruited to the site to proliferate, produce collagen, and effectively “wall off” the antigen. A granuloma requires a minimum of one to two weeks to form.

Auto-immunity In auto-immunity, as with hypersensitivity, the immune system is stimulated by specific responses that are pathogenic, and both tend to have a genetic component that predisposes some individuals more than others. However, as is the case with hypersensitivity, the adverse immune response of drug-induced autoimmunity is not restricted to the drug itself, but also involves a response to self antigens. Autoimmune responses directed against normal components of the body may consist of antibody-driven humoral responses and/or cell-mediated, delayed-type hypersensitivity responses. T cells can react directly against specific target organs, or B cells can secrete autoantibodies that target “self.” Autoimmunity may occur spontaneously as the result of a loss of regulatory controls that initiate or

106

Safety Evaluation of Medical Devices

suppress normal immunity, causing the immune system to produce lymphocytes reactive against its own cells and macromolecules such as DNA, RNA, or erythrocytes. Although autoantibodies are often associated with autoimmune reactions, they are not necessarily indicative of autoimmunity (Russel, 1981). Antinuclear antibodies can occur normally with aging in some healthy women without autoimmune disease, and all individuals have B cells with the potential of reacting with self antigens through Ig receptors (Dighiero et al., 1983). The presence of an antibody titer to a particular immunogen indicates that haptenization of serum albumin has occurred as part of a normal immune response. However, if cells are stimulated to proliferate and secrete autoantibodies directed against a specific cell or cellular component, a pathological response may result. The tissue damage associated with autoimmune disease is usually a consequence of type II or III hypersensitivity reactions that result in the deposition of antibody–antigen complexes. Several diseases have been associated with the production of autoantibodies against various tissues. For example, an autoimmune form of hemolytic anemia can occur if the antibodies are directed against erythrocytes. Similarly, antibodies that react with acetylcholine receptors may cause myasthenia gravis, those directed against glomerular basement membranes may cause Goodpasture’s syndrome, and those that target the liver may cause hepatitis. Other forms of organ-specific autoimmunity include autoimmune thyroiditis (as seen with amiodarone) and juvenile diabetes mellitus, which result from autoantibodies directed against the tissue-specific antigens thyroglobulin and cytoplasmic components of pancreatic islet cells, respectively. In contrast, systemic autoimmune diseases may occur if the autoantibodies are directed against an antigen that is ubiquitous throughout the body, such as DNA or RNA. For example, systemic lupus erythematosus (SLE) occurs as the result of autoimmunity to nuclear antigens that form immune complexes in the walls of blood vessels and basement membranes of tissues throughout the body. The etiology of renal autoimmunity is not well established and is confounded by factors such as age, sex and nutritional state, as well as genetic influences on pharmacological and immune susceptibility. Unlike idiopathic autoimmunity, which is progressive or characterized by an alternating series of relapses and remissions, drug-induced autoimmunity is thought to subside after the drug is discontinued. However, this is not certain since a major determining factor for diagnosis of a drug-related disorder is dependent on the observation of remission upon withdrawal of the drug (Bigazzi, 1988). One possible mechanism for xenobiotic-induced autoimmunity involves xenobiotic binding to autologous molecules, which then appear foreign to the immuno-surveillance system. If a self antigen is chemically altered, a specific T helper (Th) cell may see it as foreign and react to the altered antigenic determinant portion, allowing an autoreactive B cell to react to the unaltered hapten. This interaction results in a carrier–hapten bridge between the specific Th and autoreactive B cell, bringing them together for subsequent production of auto-antibodies specific to the self antigen that was chemically altered (Weigle, 1980). Conversely, a xenobiotic may alter B cells directly, including those that are autoreactive. Thus the altered B cells may react to self antigens independent from Th-cell recognition and in a nontissue-specific manner. Another possible mechanism is that the xenobiotic may stimulate nonspecific mitogenicity of B cells. This could

result in a polyclonal activation of B cells with subsequent production of auto-antibodies. Alternatively, the xenobiotic may stimulate mitogenicity of T cells that recognize self, which in turn activate B-cell production of antibodies in response to “self” molecules. There is also evidence to suggest that anti-DNA auto-antibodies may originate from somatic mutations in lymphocyte precursors with antibacterial or antiviral specificity. For example, a single amino acid substitution resulting from a mutation in a monoclonal antibody to polyphorlcholine was shown to result in a loss of the original specificity and an acquisition of DNA reactivity similar to that observed for anti-DNA antibodies in SLE (Talal, 1987). The mechanism of autoimmunity may also entail interaction with MHC structures determined by the HLA alleles. Individuals carrying certain HLA alleles have been shown to be predisposed to certain autoimmune diseases, which may account in part for the genetic variability of autoimmunity. In addition, metabolites of a particular drug may vary between individuals to confound the development of drug-induced autoimmunity. Dendritic cells, such as the Langerhans cells of the skin and B lymphocytes that function to present antigens to Th cells, express class-II MHC structures. Although the exact involvement of these MHC structures is unknown, Gleichmann et al. (1989) have theorized that self antigens rendered foreign by drugs such as D-penicillamine may be presented to Th cells by MHC class-II structures. An alternate hypothesis is that the drug or a metabolite may alter MHC class-II structures on B cells, making them appear foreign to Th cells. A number of different drugs have been shown to induce autoimmunity in susceptible individuals A syndrome similar to that of SLE was described in a patient administered sulfadizine in 1945 by Hoffman (see Bigazzi, 1985). Sulfonamides were one of the first classes of drugs identified to induce an autoimmune response, while to date, more than 50 other drugs have been associated with a similar syndrome. Autoantibodies to red blood cells and autoimmune hemolytic anemia have been observed in patients treated with numerous drugs, including procainamide, chlorpropamide, captopril, cefalexin, penicillin, and methyldopa (Logue et al., 1970; Kleinman et al., 1984). Hydralazine- and procainamide-induced autoantibodies may also result in SLE. Approximately 20% of patients administered methyldopa for several weeks for the treatment of essential hypertension developed a dose-related titer and incidence of autoantibodies to erythrocytes, 1% of which presented with hemolytic anemia. Methyldopa does not appear to act as a hapten but appears to act by modifying erythrocyte surface antigens. IgG auto-antibodies then develop against the modified erythrocytes. Some metals that are used therapeutically have also been shown to induce autoimmune responses. Gold salts used to treat arthritis may induce formation of antiglomerular basement membrane antibodies, which may lead to glomerulonephritis similar to that seen in Goodpasture’s disease (see type II hypersensitivity). Since gold is not observed at the site of the lesions (Druet et al., 1982) it has been hypothesized that the metal elicits an antiself response. Lithium, used to treat manic-depression, is thought to induce autoantibodies against thyroglobulin, which in some patients results in hypothyroidism. In studies with rats, levels of antibodies to thyroglobulin were shown to increase significantly in lithium-treated rats compared to controls immediately after immunization with thyroglobulin; however,

Chapter 9:

rats that were not immunized with throglobulin did not produce circulation antithroglobulin antibodies upon receiving lithium, and there was no effect of lithium on lymphocytic infiltration of the thyroid in either group (Hassman et al., 1985). In addition, silicone-containing medical devices, particularly breast prostheses, have been reported to cause serum-sickness-like reactions, scleroderma-like lesions, and an SLE-like disease termed human adjuvant disease (Kumagai et al., 1984; Guillaume et al, 1984). Some patients may also present with granulomas and autoantibodies. Human adjuvant disease is a connective tissue or auto-immune disease similar to that of adjuvant arthritis in rats and rheumatoid arthritis in humans. Auto-immune disease-like symptoms usually develop 2–5 yr after implantation in a small percentage of people that receive implants, which may indicate that there is a genetic predisposition similar to that for SLE. An early hypothesis is that the prosthesis or injected silicone plays an adjuvant role by enhancing the immune response through increased macrophage and T-cell helper function. There is currently controversy as to whether silicone, as a foreign body, induces a nonspecific inflammation reaction, a specific cell-mediated immunological reaction, or no reaction at all. However, there is strong support to indicate that silicone microparticles can act as haptens to produce a delayed hypersensitivity reaction in a genetically susceptible population of people.

EVALUATION OF THE IMMUNE SYSTEM Regulatory Positions Since the last edition of this book the pharmaceutical and medical device industries have come to have specific regulatory guidances requiring routine evaluation of the toxicological profile of the xenobiotic (drugs and devices) in routine preclinical safety testing. The chemical industry has been a proponent of using a battery of assays to assess chemicalinduced immunotoxicity, hence guidelines for a two-tiered screen approach have been proposed by the National Toxicology Program (NTP) (Luster et al., 1988). This strategy, which was developed for nontherapeutic chemicals and environmental contaminants that have different safety standards, does not address some of the safety issues and test strategy issues that are unique to pharmaceuticals. ICH and ISO (and by reference the FDA) have promulgated drugs (ICH) and devices (ISO 10993-20) a similar two-leveled approach (Hinton, 1992) for assessing immunotoxicity of food colors and additives. In all of these testing schemes, the initial tier generally includes a fundamental histopathologic assessment of the major components of the immune system. Additional tiers are then added to more precisely evaluate the functionality of the components that appeared to be adversely affected in the first tier of tests. These test strategies are primarily geared toward the detection of chemicalinduced immunosuppression, thus the effectiveness of these test schemes for detecting immunostimulation has not yet been determined (Spreafico, 1988). ISO defines the first tier of assays (Table 9.8) to include an assessment of immunopathology: humoral, cell-mediated, and nonspecific immunity such as natural killer cell activity. The second tier (Table 9.9) includes a more comprehensive battery that should be used once functional changes are observed in the Tier I assays. The Tier II assays focus on mechanisms of immunotoxicity such as depletion of specific cell subsets by flow cytometry analysis or evaluation of secondary immune responses by examining IgG response.

Table 9.8

Immunotoxicology

107

Tier I Screen

Parameter

Procedures

Immunopathology

Routine hematology—complete and differential count; routine toxicology information—weights of body, immune organs (spleen and thymus), liver, and kidney; histopathology of immune organs. LPS (lipopolysaccharide) mitogen response or F(ab)2 mitogenic response; enumeration of plaques by IgM antibody-forming cells to a T-dependent antigen (sheep red blood cells; serum IgM concentration Lymphocyte mitogenic response to concanavalin A and mixed lymphocyte response to allogeneic lymphocytes; local lymph node assay. Natural killer cell activity

Humoral-mediated immunity

Cell-mediated immunity Nonspecific immunity

Sources: Adapted from Luster et al. (1988) and Vos et al. (1989).

Cell-mediated immunity is assessed through a functional assay that looks at the ability of cytotoxic T cells to kill target cells, and nonspecific immunity is evaluated by examining various function of macrophages: (1) the ability to phagocytize inert fluorescent beads or radiolabeled chicken erythrocytes and (2) the ability to produce cytokines such as IL-1 or macrophage activation factor. The ultimate immune test would be to examine the effects of xenobiotics on the intact animal’s response to challenge by viral, bacterial, or parasitic pathogens, or neoplastic cells. The ability of the immune system to compensate or, conversely, its inability to compensate for loss or inhibition of its components is fully examined through host resistance mechanisms. This tiered test approach has been validated with 50 selected compounds, and results from these studies have shown that the use of only two or three immune tests are sufficient to predict known immunotoxic compounds in rodents with a >90% concordance (Luster et al., 1992a, b). Specifically, the use of Table 9.9

Tier II Screen

Parameter

Procedures

Immunopathology

Enumeration of T and B cells and subsets; immunocytochemistry of lymphoid tissues; inumeration of cell types and numbers in the bone marrow. Enumeration of secondary antibody (IgG) response to sheep red blood cells Cytotoxic T lymphocyte killing; delayed type hypersensitivity response; mouse ear swelling test (MEST; Gad et al., 1986); guinea pig maximization test (Magnusson and Kligman, 1969). Macrophage function—in vitro phagocytosis of fluorescent covaspheres, killing of Listeria monocytogenes or of tumor cells [basal and activated by macrophage activating factor (MAF)]. Bacterial models—Listeria monocytogenes (mortality or spleen clearance); Streptococcus species (mortality); Viral models—influenza (mortality); Parasitic models—Plasmodium yoelii (parasitemia) or Trichinella spiralis (muscle larvae counts and worm expulsion); Syngeneic tumor models—PYB6 sarcoma (tumor incidence); B16F10 melanoma (lung burden).

Humoral-mediated immunity Cell-mediated immunity

Nonspecific immunity

Host resistance

Sources: Adapted from Luster et al. (1988) and Vos et al. (1989), unless otherwise indicated.

108

Safety Evaluation of Medical Devices

either a humoral response assay for plaque-forming colonies (PFC response) or determination of surface marker expression, in combination with almost any other parameter, significantly increased the ability to predict immunotoxicity when compared to the predictivity of any assay alone. The FDA guidelines for immunotoxicity testing of food additives start with a Type 1 battery of tests. Type 1 tests can be derived from the routine measurements and examinations performed in short-term and subchronic rodent toxicity studies, since they do not require any perturbation of the test animals (immunization or challenge with infectious agents). These measurements include hematology and serum chemistry profiles, routine histopathologic examinations of immune-associated organs and tissues, and organ and body weight measurements, including thymus and spleen. If a compound produces any primary indicators of immunotoxicity from these measurements, more definitive immunotoxicity tests, such as those indicated in the preceding paragraph, may be recommended on a case-by-case basis. The following is a brief explanation of some of the indicators that may be used to trigger additional definitive testing and a description of some of the most commonly used assays to assess humoral, cell-mediated, or nonspecific immune dysfunction, which are common to most immunotoxicology test strategies.

CDRH Testing Framework The CDRH draft document (1997) actually sets forth a concise and step-wise approach to evaluating the potential immunotoxicity risks of devices (Figure 9.2 and Table 9.10). If the process in Figure 9.2 identifies a potential for hazard then the tests specified in Table 9.10 are employed to evaluate those risks.

IMMUNOPATHOLOGIC ASSESSMENTS Various general toxicological and histopathologic evaluations of the immune system can be made as part of routine

preclinical safety testing to obtain a preliminary assessment of potential drug-related effects on the immune system. At necropsy, various immunological organs of the immune system such as thymus, spleen, and lymph nodes are typically observed for gross abnormalities and weighed in order to detect decreased or increased cellularity. Bone marrow and peripheral blood samples are also taken to evaluate abnormal types and/or frequencies of the various cellular components. Organ and body weights. Changes in absolute weight, organ-to-body weight ratios, and organ-to-brain weight ratios of tissues such as thymus and spleen are useful general indicators of potential immunotoxicity. However, these measures are nonspecific for immunotoxicity since they may also reflect general toxicity and effects on endocrine function that can indirectly affect the immune system. Hematology. Hemacytometers or electronic cell counters can be used to assess the numbers of lymphocytes, neutrophils, monocytes, basophils, and eosinophils in the peripheral blood, while changes in relative ratios of the various cell types can be assessed by microscopic differential evaluation. Similar evaluations can be performed with bone marrow aspirates, where changes may reflect immunotoxicity to the pluripotent stem cells and newly developing lymphoid precursor cells. Potential hematological indicators of immunotoxicity include altered white blood cell counts or differential ratios, lymphocytosis, lymphopenia, or eosinophilia. Changes in any of these parameters can be followed up with more sophisticated flow cytometric analyses or immunostaining techniques that are useful for phenotyping the various types of lymphocytes (B cell, T cell) and the T-cell subsets (CD4+ and CD8+) on the basis of unique surface markers. Decreases or increases in the percentages of any of the cell populations relative to controls, or in the ratios of B cells/T cells, or CD4+/CD8+ cells may be indicators of immunotoxicity. Clinical chemistry. Nonspecific clinical chemistry indicators of potential immune dysfunction include changes in serum protein levels in conjunction with changes in the albumin-to-globulin (A/G) ratio. Immunoelectrophoretic

Device contacts the body directly or indirectly? Yes

No

Device contains potentially immunotoxic material?

No

No additional immunotoxicity testing is required.

Yes Device material exactly the same as in a legally marketed device with nontoxicity/long-time use.

Yes

Same body contact as in a legally marketed device. No

No Immunotoxicity data provided based on FDA immunotoxicity testing framework? No Immunotoxicity test data required.

Figure 9.2

CDRH Risk Assessment Immunotoxicity Flow Chart.

Yes

Yes

Chapter 9:

analysis of serum proteins can then be performed to quantify the relative percentages of albumin and the α-, β-, and γ-globulin fractions. To perform these assays, a drop of serum (antigen) is placed into a well cut in a gel, then the gel is subjected to electrophoresis so that each molecule in the serum moves in the electric field according to its charge. This separation is then exposed to specific antiserum, which is placed in a trough cut parallel to the direction in which the components have moved. By passive diffusion, the antibody reaches the electrophoretically separated antigen and reacts to form Ag–Ab complexes. The γ-globulin fractions can be separated and further quantified for the relative proportions of IgG, IgM, IgA, and IgE using similar techniques. Serum concentrations of immunoglobulin classes and subclasses can also be measured using various techniques such as radioimmunoassays (RIAs) or enzyme-linked immunosorbent assays (ELISAs). In the ELISA, antigens specific for each class of immunoglobulin can be adsorbed onto the surfaces for microtiter plates. To determine the quantity of each antibody in a test sample, an aliquot of antiserum is allowed to react with the adsorbed antigens. Unreacted molecules are rinsed off and an enzyme-linked anti-Ig is then added to each well. Next, substrate is added and the amount of color that develops is quantified using a spectrophotometric device. The amount of antibody can then be extrapolated from standard curves since the amount of color is proportional to the amount of enzyme-linked antibody that reacts. Variations in levels of a given antibody may indicate the decreased ability of B cells or decreased numbers of B cells producing that antibody. In addition, serum autoantibodies to DNA, mitochondria, and parietal cells, can be used to assess autoimmunity. Serum cytokines (IL-1, IL-2, and γ-interferon) can also be evaluated using immunochemical assays to evaluate macrophage, lymphocyte, and lymphokine activity; prostaglandin E2 can also be measured to evaluate macrophage function. CH50 determinations can be used to analyze the total serum complement and are useful for monitoring immune complex diseases (Sullivan, 1989); activation of complement (Table 9.11) in the presence of auto-antibodies is indicative of immune complex diseases and auto-immunity. The various components of the complement system (C3, C4) can also be measured to assess the integrity of the system. For instance, low serum concentrations of C3 and C4, with a concomitant decrease in CH50, may indicate activation of complement, while a low C4 alone is a sensitive indicator of reduced activation of the complement system. Since C3 is used as an alternate complement pathway, it usually measures high. Therefore, a low C3 with a normal C4 may indicate an alternate pathway of activation. Histopathology. Histopathologic abnormalities can be found in lymphoid tissues during gross and routine microscopic evaluations of the spleen, lymph nodes, thymus, bone marrow, and gut-associated lymphoid tissues such as Peyer’s patches and mesenteric lymph nodes. Microscopic evaluations should include descriptive qualitative changes such as types of cells, densities of cell populations, proliferation in known T- and B-cells areas (e.g., germinal centers), relative numbers of follicles and germinal centers (immune activation), and the appearance of atrophy or necrosis. In addition, unusual findings such as granulomas and scattered, focal mononuclear cell infiltrates in nonlymphoid tissues may be observed as indicators of chronic hypersensitivity or autoimmunity. A complete histopathologic evaluation should also include a quantitative assessment of

Immunotoxicology

109

cellularity through direct counts of each cell type in the various lymphoid tissues. In addition, changes in cellularity of the spleen can be more precisely quantitated from routine H & E (hematoxylin and eosin) sections using morphometric analysis of the germinal centers (B cells) and periarteriolar lymphocyte sheath (T cells). Similar morphometric measurements can be made of the relative areas of the cortex and medulla of the thymus. If changes in cellularity are apparent from routinely stained histopathology sections, special immunostaining (immunoperoxidase or immunofluorescence) of B cells in the spleen and lymph nodes using polyclonal antibodies to IgG, or immunostaining of the T cells and their subsets in the spleen using mono-polyclonal antibodies to their specific surface markers, can be used to further characterize changes in cellularity. Numerous physiological and environmental factors such as age, stress, nutritional deficiency, and infections may affect the immune system (Sullivan, 1989). Thus, adverse findings in animal studies may reflect these indirect immunotoxic effects rather than the direct immunotoxic potential of a chemical or drug. Indirect immunotoxic effects may be assessed through histopathologic evaluations of endocrine organs such as the adrenals and pituitary. It is also well known that the functional reserves of the immune system can allow biologically significant, immunotoxic insults to occur without the appearance of morphological changes. In addition, there is some built-in redundancy in the system in that several mechanisms may produce the same outcome. For instance, cytotoxic T cells may alone be sufficient to protect the organism against a bacterial infection; however, the body will also produce antibodies for future protection. Thus, if one mechanism is insufficient to fight off infection, the second mechanism can serve as a backup. Because of this functional reserve, adverse effects may remain subclinical until the organism is subjected to undue stress or subsequent challenge (Bloom et al., 1987). Therefore, routine immunopathologic assessments as part of standard preclinical toxicity tests may not be sufficient to detect all immunotoxins. Although changes detected in routine toxicological and pathological evaluations are nonspecific, and of undetermined biological significance to the test animal, they can be invaluable as flags for triggering additional testing.

Humoral Immunity As described previously, the humoral immune response results in the proliferation, activation, and subsequent production of antibodies by B cells following antigenic exposure and stimulation. The functionality and interplay between the three primary types of immune cells (macrophage, B cells, and T cells) required to elicit a humoral response can be assessed through various in vitro assays using cells from the peripheral blood or lymphoid tissues. Antibody plaque-forming cell (PFC) assay. The number of B cells producing antibody (PFC) to a T-dependent antigen such as sheep red blood cells (SRBCs) can be assessed in vitro following in vivo exposure to the test article and antigen (ex vivo tests). The PFC response to a T-dependent antigen is included as a Tier I test by the NTP since it appears to be the most commonly affected functional parameter of exposure to immunosuppressants. However, this test is designated as a Type 2 test in the FDA Redbook since it requires an in vivo immunization of the animals with antigen and, thus, cannot be evaluated as part of an initial toxicity screen.

110

Safety Evaluation of Medical Devices

Table 9.10

CDRH Immunotoxicity Testing Matrix Immunological effects

Body contact

Contact duration

1

Surface device Skin

<24 h (A) 24–30 d (B) >30 d (C)

pmb pmb pmb

2

3

4

5

Mucosal membrane

A B C

pmb pmb pmb

pmb pmb pmb

mb mb

Breached/compromised Surface

mb

mb

A B C

pmb pmb pmb

pmb pmb pmb

mb mb

mb mb

mb mb

External comm. device Blood path, direct and indirect

A B C

pmb pmb pmb

pmb pmb pmb

mb mb

pmb pmb

pmb mb

Tissue/bone/dentin Communicating

A B C

pmb pmb pmb

pmb all all

mb mb

pmb pmb

mb mb

Implant devices

A B C

pmb pmb pmb

pmb all all

mb mb mb

pmb pmb

mb mb

P: Plastics/polymers B: Biological material M: Metals

1: Hypersensitivity 2: Inflammation 3: Immunosuppression

Although this assay requires that B cells be fully competent in secreting antibodies, T cells and macrophage cells are also essential for the proper functioning of humoral immunity. However, this assay is nonspecific in that it cannot determine which cell type(s) is responsible for dysfunction.

Table 9.11 Response factors Complement fixation

The Complement System Origin

Characteristics/functions

Serum

Critical component of humoral immune response leading to lysis of cell membranes, chemotaxis, and phagocytosis. Binds with IgG or IgM on membrane of the target cell to initiate activation of complement cascade Activated by C1; act together as a complex to activate C3; exposes a membrane site recognizable to granulocytes and macrophages, resulting in opsonization. C3 binds to C42 complex to form C423; C5 binds to C423 to form C4235; provides sites for C6 & C7. C6 & C7 bind to C5 site to result in C567. One molecule of C567 binds with C8 to result in C5678. Up to six C9 molecules can bind with C5678 to trigger lysis by disrupting the lipid layer of the cell membrane.

C1

C4 and C2

C3 and C5

C6 and C7 C8 C9

Liver, macrophage

4: Immunostimulation 5: Autoimmunity

Macrophage cells are needed to process antigen and produce IL-1. T cells are needed for several functions including antigen recognition of surface membrane proteins and B-cell maturation through the production of various lymphokines that stimulate growth and differentiation. SRBCs are most commonly used as the T-dependent antigen, although T-cell independent antigens may also be useful to rule out T helper dysfunction as a cause of immunodysfunction. The PFC assay has evolved from methodology originally developed as a hemolytic plaque assay (or Jerne plaque assay) by Nils Jerne to quantitate the number of antibodyforming cells in a cell suspension plated with RBCs onto agar plates (Jerne and Nordin, 1963). In its present form, animals are treated in vivo with the test compound, immunized with approximately 5 × 108 SRBCs administered intravenously within two to three days post-treatment, and then sacrificed four days (IgM) or six days (IgG) later. Antibody-producing spleen cell suspensions are then mixed in vitro with SRBCs, placed onto covered slides, and incubated for a few hours in the presence of complement. During incubation, antibody diffuses from the anti-SRBC-producing cells and forms Ag–Ab complexes on the surfaces of nearby SRBCs, resulting in the formation of small clear plaques on the slide. Plaques are then counted and expressed as PFCs/106 spleen cells. A dose-related reduction in PFCs is indicative of immunosuppression. B-cell lymphoproliferation response. The NTP has classified this assay as a Tier I test since mitogenesis can be performed easily in tandem with other tests to provide an assessment of the proliferative capacity of the cells (Luster et al., 1988). Since this assay is performed ex vivo with peripheral blood (or spleen) and is well characterized for use in various animal species, it has also been included as an Expanded Type 1 test in the revised Redbook. The proliferation of peripheral blood or splenic B cells following stimulation with lipopolysaccharide (LPS) or

Chapter 9:

other mitogens (pokeweed mitogen extract) is another measurement of humoral immunity. LPS (a bacterial lipopolysaccharide) is a B-cell-specific mitogen that stimulates polyclonal proliferation (mitosis) as part of the natural sequence of antigen recognition, activation, and clonal expansion. The mitogen does not interact with just one particular antigen-specific clone, but with all cells bearing the carbohydrate surface marker for which it is specific. Since mitogens are both polyclonal and polyfunctional, they can stimulate a wider spectrum of antigenic determinants than antigens, which can only stimulate a low number (10−6) of specific cells. In this assay, lymphocytes from animals treated in vivo are cultured in vitro in microtiter plates in the presence of tritiated [3H]thymidine (or uridine) using a range of at least three concentrations of mitogen to optimize the response. Lymphocytes can be obtained aseptically from peripheral blood or from single cell suspensions of spleen cells that are prepared by pushing the tissue through sterile gauze or 60-mesh wire screens. A decrease in DNA synthesis (incorporation of 3H) as compared to the unexposed cells of control animals may indicate that the B cells were unable to respond to antigenic stimulation. Alternative methodology employs an 18–20 hr incubation with 125I-labeled iododeoxyuridine ([125I]IUdR) and fluorodeoxyuridine (FUdR) ((White et al., 1985). After incubation, the cells are collected onto filter disks and then counted with a gamma counter. Assays such as this that use polyclonal mitogens for activation may not be as sensitive as specific antigen-driven systems (Luster et al., 1988). In addition, suppression of the mitogen response does not always correlate with the PFC response. Since mitogenesis represents only a small aspect of B-cell function and maturation, this end point is not sensitive to early events that may affect activation, or later events that may affect differentiation of B cells into antibody-secreting cells (Klaus and Hawrylowicz, 1984).

Cell-Mediated Immunity T-cell lymphoproliferation response. This assay is analogous to the B-cell lymphoproliferative response assay described above. Thus, this assay is also classified as a Tier I test by the NTP and as an Expanded Type 1 test in the revised draft of the Redbook. T cells from the peripheral blood or spleen undergo blastogenesis and proliferation in response to specific antigens that evoke a cell-mediated immune response. T-cell proliferation is assessed using T-cell-specific mitogens such as the plant lectins, concanavalin A (Con A), and phytohemagglutinin (PHA) or T-cell-specific antigens (i.e., tuberin, Listeria). Uptake of 3H as an indicator of DNA synthesis is used as described above for evaluating B-cell proliferation. T-cell mitogens do not just stimulate synthesis of DNA but, in fact, they also stimulate the expression of cell specific function. For instance, Con A can trigger the expression of T helper, suppressor, and cytotoxic effector cells, and either mitogen may induce the expression (or reexpression of memory cells) of the differentiated function (Clark et al., 1983). Since cell populations responsive to Con A are thought to be relatively immature compared to those that are stimulated with PHA, the parallel usage of both mitogens may be useful for distinguishing the affected subset (Tabo and Paul, 1973). A secondary response to T-cell antigens such as purified protein derivative of tuberculin (PPD) or tetanus toxoid can also be assessed.

Immunotoxicology

111

Mixed lymphocyte response (MLR) assay. This assay has been shown to be sensitive for the detection of chemicalinduced immunosuppression and is a recommended Tier I assay by the NTP (Luster et al., 1988). In addition, it has been shown to be predictive of host response to transplantation and of general immunocompetence (Harmon et al., 1982). The mixed lymphocyte response assay assesses the ability of T cells to recognize foreign antigens on allogenic lymphocytes and, thus, is an indirect measure of the cellmediated ability to recognize graft or tumor cells as foreign. Responder lymphocytes from animals treated in vivo with the test compound are mixed with allogeneic stimulator lymphocytes that have been treated in vitro with mitomycin C or irradiated to render them unable to respond (Bach and Voynow, 1966). Both cell types are cultured in vitro for three to five days, then incubated with 3H for an additional 6 hr. Once the radiolabel is incorporated into the DNA of the responding cells, the DNA is extracted and the amount of radioactive label is measured to quantitate proliferation of the responder cells of drug-treated animals compared to those of the controls. Cytotoxic T lymphocyte (CTL) mediated assay. This assay is similar to the MLR assay and can be performed in parallel or as a Tier II follow-up to the MLR assay. The CTL assay ascertains the ability of cytotoxic T cells to lyse an allogeneic target cell or the specific target cell type with which they were immunized. In general, the cytolytic response of activated effector cells is assessed by measuring the amount of radioactivity (51Cr) that is released from the target cell. When performed in conjunction with the MLR assay, lymphoid cells of the two strains are cultured together in vitro as described above; however, 51Cr is added to the culture after four to five days (instead of 3H). Both responder and target cells are labeled with the 51Cr, which is taken up rapidly by the cells through passive diffusion but is released slowly as long as the cell membrane is intact. Furthermore, since chromium reduced from Cr6+ to Cr3+ enters the cells at a much slower rate than Cr6+, the 51Cr released from the damaged target cells is not significantly reincorporated into undamaged cells (Clark, 1983), which would reduce the sensitivity of the assay. Thus, the amount of chromium released into the medium and recovered in the supernatant of the mixture of the cells is directly proportionate to the extent of lysis of the target cells by the sensitized responder cells. In a capillary tube assay developed in 1962 by George and Vaughan, the inhibition of migration of macrophage cells can be used to access normal T-cell function (see Clark, 1983). T cells are obtained from the peripheral blood of animals treated in vivo with a test article and injected with an antigen(e.g., tuberculin). If these T cells are functioning normally, they should release migration inhibition factor (MIF). As a consequence, the macrophages, which generally show a propensity for migration upon stimulation with the antigen, should show a MIG-induced reduction in migratory behavior. Delayed-type hypersensitivity (DTH) response. The DTH response assay is considered to be a comprehensive Tier II assay for cell-mediated immunity by the NTP. To express a DTH inflammatory response, the immune system must be capable of recognizing and processing antigen, blastogenesis and proliferation of T cells, migration of memory T cells to the challenge site of exposure to antigen, and subsequent production of inflammatory mediators and lymphokines that elicit the inflammatory response. Thus, by measuring a DTH response to an antigen, these assays assess the functional status of both the afferent

112

Safety Evaluation of Medical Devices

(antigen recognition and processing) and efferent (lymphokine production) arms of cellular immunity. Various antigens have been used for assessing DTH, including keyhole limpet hemocyanin (KLH), oxazolone, dinitrochlorobenzene, and sheep red blood cells (SRBCs) (Vos, 1977; Godfrey and Gell, 1978; Luster et al., 1988). In one such assay described by White et al. (1985), mice previously treated with the test article are sensitized to SRBCs by inoculation of SRBCs into the hind footpad and four days later challenged in the same footpad. Seventeen hours following the challenge, they are injected intravenously with 25I-labeled human serum albumin (HSA), then sacrificed 2 hr later. Both hind feet are then radioassayed in a gamma counter (the second foot serves as a control for background infiltration of the label). With a normal functioning cell-mediated response,125I-labeled HSA will extravasate into the edematous area produced by the DTH response (Paranjpe and Boone, 1972). In general, a decrease in the extravasation of 125 I-labeled HSA is indicative of immunosuppression of the efferent arm of the cell-mediated immune system. To assess specifically the afferent arm of the DTH response, the proliferation of the popliteal lymph node cells to SRBCs can also be measured (White et al., 1985). As described above, mice treated with the test article are sensitized to SRBCs by inoculation of SRBCs into the hind footpad. However, 1.5 hr later they are challenged intraperitoneally with FUdR and 2 hr later they are administered [125I]IUdR intravenously (instead of 125I-labeled HSA). Mice are sacrificed 24 hr after challenge and both popliteal lymph nodes are removed and counted in a gamma counter. Similar assays for DTH have been traditionally performed with the antigen Mycobacterium tuberculosis, which preferentially elicits a cell-mediated response. In this assay a small amount of antigen contained in the supernatant fluid from the medium in which the pathogen was grown is injected into the footpad. Upon challenge, a visible and palpable lump should appear by 48 hr. The amount of swelling is then measured and compared with the footpad that did not receive the challenge. Alternatively, methods used by NTP employ a modified 125I-labeled uridine (UdR) technique to measure the monocyte influx at the challenge site (ear) injected with keyhole limpet hemocyanin antigen. This assay has been shown to correlate well with decreased resistance to infectious disease (Luster et al., 1988). However, one should note that regardless of which technique is used, antiinflammatory drugs may produce false-positive results in this type of assay.

Nonspecific Immunity Natural killer cell assays. This assay is a Tier I test for nonspecific immunity in the NTP testing scheme (Luster et al., 1988) and is proposed as an additional Type 1 test in the draft Redbook. Natural killer (NK) cells, like cytotoxic T cells, have the ability to attack and destroy tumor cells or virus-infected cells. However, unlike T cells, they are not antigen specific, do not have unique, clonally distributed receptors, and do not undergo clonal selection. In in vitro or ex vivo tests, target cells (e.g., YAC-1 tumor cells) are radiolabeled in vitro or in vivo with 51Cr and incubated in vitro with effector NK cells from the spleens of animals that had been treated with a xenobiotic. This assay can be run in microtiter plates over the range of various ratios of effector/target cells. Cytotoxic activity is then measured by the amount of radioactivity released from the damaged tumor cells as was previously

described for cytotoxic T cells. This assay can also be performed in vivo, where YAC-1 cells labeled with [125I]IUdR are injected directly into mice and NK cell activity is correlated with its level of radioactivity (Riccardi et al., 1979). Immunotoxicity observed as reduced NK cell activity is correlated with increased tumorigenesis and infectivity. Macrophage function. Several assays are available to measure various aspects of macrophage function, including quantitation of resident peritoneal cells, antigen presentation, cytokine production, phagocytosis, intracellular production of oxygen free radicals (used to kill foreign bodies) and direct tumor-killing potential. Techniques for quantitation of peritoneal cells and functional assays for phagocytic ability are classified as comprehensive Tier II tests by the NTP and as additional Type 1 tests in the draft Redbook. Macrophage cells and other polymorphonuclear cells (PMNs) contribute to the first-line defense of non-specific immunity through their ability to phagocytize foreign materials, including pathogens, tumor cells, and fibers (e.g., silica, asbestos). Xenobiotics can affect macrophage function direct toxicity to macrophages or by modulating their ability to become activated. Differential counts of resident peritoneal cells can be made as rapid, preliminary assessment of macrophage function for xenobiotics that are not administered parenterally. Numerous in vitro assays can be employed to assess common function of macrophages and PMNs including adherence to glass, migration inhibition, phagocytosis, respiratory activity (chemiluminescent assays or nitroblue tetrazolium), and target cell killing. In one such assay, the chemotactic response to soluble attractants is evaluated using a Boyden chamber with two compartments that are separated by a filter. Macrophage cells or PMNs from treated animals are placed in one side and a chemotactic agent in the other. Chemotaxis is then quantified by counting the number of cells that pass through the filter. In another assay, the ability of the macrophages to phagocytize foreign materials can be evaluated by adding fluorescent latex beads to cultures containing macrophage cells, then determining the proportion of cells that have phagocytized the beads using a fluorescent microscope or by flow cytometry (Duke et al., 1985). Similar functions can be evaluated by incubating the cells with known amounts of bacteria. The cells are then removed by filtration or centrifugation, the remaining fluid is plated onto bacterial nutrient agar, and, after a few days of incubation, the bacterial colonies are counted. Furthermore, the efficiency of the cells to kill the bacteria once phagocytized can be assayed by lysing the cells and plating the lysate onto bacterial agar. Various in vivo assessments of macrophage function have also been used. For example, peritoneal exudate cell (PEC) recruitment can be assessed using eliciting agents such as Corynebacterium parvum, MVE-2, or thioglycolate (Dean et al., 1984). In one such assay (White et al., 1985), mice are injected intraperitoneally with thioglycolate, sacrificed five days later, and the peritoneal cavity is flushed with culture medium. The cell suspension is then counted, the cell concentration is adjusted to a known density (2 × 105 mL−1), and the cells are cultured for 1 hr in 24-well culture dishes. Adherent cells are then washed with medium and aliquots of 51Cr-labeled SRBCs that were opsonized with mouse IgG are added to each well and incubated for various times. This same system can be used to assess adherence to and chemotaxis of the PECs (Laskin et al., 1981). Phagocytosis can also be evaluated in vivo by

Chapter 9:

measuring the clearance of injected particles from the circulation and the accumulation of the particles in lymphatic tissues such as the spleen. Mast cell/basophil function. The function of mast cells and basophils to degranulate can be evaluated using a passive cutaneous anaphylaxis test (Cromwell et al., 1986). Serum containing specific anaphylactic (IgE) antibodies from donor animals previously exposed to a known antigen is first administered by intradermal (or subcutaneous) injection into unexposed host animals. After a sufficient latency period to allow binding of the donor IgE to the host tissue mast cells, the animals are administered a second intravenous injection of the antigen. The anaphylactic antibodies present in the serum will stimulate normally functioning mast cells to degranulate (release histamines) and produce a marked inflammatory response. Using similar in vitro assays with mast cells and basophils, the quantities of histamines that are released from the cells can be measured directly in the culture medium.

Host-Resistance Assays Host-resistance assays can be used to assess the overall immunocompetence of the humoral or cell-mediated immune systems of the test animal (host) to fend off infection with pathogenic microbes, or to resist tumorigenesis and metastasis. These assays are performed entirely in vivo and are dependent on all of the various components of the immune system to be functioning properly. Thus, these assays may be considered to be more biologically relevant than in vitro tests that only assess the function of cells from one source and of one type. Since these assays require that the animal be inoculated with a pathogen or exogenous tumor cell, they cannot be performed as part of a general preclinical toxicity assessment, and are thus classified as Type 2 tests in the revised Redbook. These assays are also included as Tier II tests by the NTP. Similar host-resistance assays are used to evaluate the immunosurveillance of spontaneous tumors, which is assessed as the capacity of the organism to reject grafted syngeneic tumors. Various animal-bearing tumor models (Pastan et al., 1986) and host-resistance models have been used to assess immunotoxicity. Several of the host-resistance assays utilize cultured tumor cell lines such as PYB6 sarcoma and B16F10 melanoma cells that are used with Fischer 344 rats. For example, the PYB6 sarcoma model uses death as an end point. In this assay, syngeneic mice are injected with the PYB6 sarcoma cells and death due to tumor is recorded daily. In another routinely used assay, animals that have been treated with a xenobiotic are injected with either B16F10 melanoma cells or Lewis lung carcinoma cells, then approximately 20 days later they are sacrificed and pulmonary tumors are measured and counted.

Hypersensitivity Type I hypersensitivity. Although there are acceptable systems for evaluating Type I (immediate) reactions following systemic exposure, there are no reliable animal models for predicting Type I reactions following dermal applications or oral administration of drug. Repeated exposure of a xenobiotic is required to produce a Type I response. A drug in the form of a hapten must covalently bind to macromolecules (proteins, nucleic acids) before it can initiate a primary antibody response. Once sensitized, even the smallest exposure to the xenobiotic can elicit a rapid, intensive IgE antibodymediated inflammatory response. With the exception of

Immunotoxicology

113

antivirals and chemotherapeutic drugs, most drugs should not be reactive with biological nucleophiles since these drugs are usually screened out as mutagens or carcinogens in preclinical safety studies. However, Type I hypersensitivity is a particular problem with biotechnology products themselves (e.g., insulin, growth hormones, interleukins), trace impurities from the producing organisms (e.g., E. coli proteins, mycelium), or the vehicles used to form emulsions (Matory et al., 1985). The production of neutralizing antibodies to recombinant DNA protein products or their contaminants may be assayed using ELISAs or IRAs. A suitable animal model used to evaluate the potential for a Type I response to protein hydrolysates is detailed in the U.S. Pharmacopoeia. This test is very sensitive for testing proteins administered by the parenteral route, but is of little value for low molecular weight drugs and those that are administered orally (Descotes and Mazue, 1987). Active systemic anaphylaxis can be assessed in guinea pigs following systemic exposure to the test compound. For dermal exposures, however, rabbits or guinea pigs must be exposed to the test article by intradermal injections and then evaluated for their ability to mount a systemic anaphylactic response. The passive cutaneous anaphylaxis test (as described above for mast cells) can also be used to assess a potential anaphylactic response to a test compound. The serum containing potential anaphylactic (IgE) antibodies from donor animals previously exposed to the test compound is first administered by intradermal (or subcutaneous) injection into unexposed host animals. After a latency period, the animals are administered an intravenous injection of the test compound together with a dye. If anaphylactic antibodies are present in the serum, the subsequent exposure to the test compound will cause a release of vasoactive amines (degranulation of mast cells), ultimately resulting in the migration of the dye to the sites of the intradermal serum injection. Types II and III hypersensitivity. No simple animal modes are currently available to assess Type II (antibodymediated cytotoxicity) hypersensitivity reactions. IgE antibodies and immune complexes in the sera of exposed animals can be assayed using ELISA or RIA techniques that require the use of specific antibodies to the drug. Type III (immune complex related disease) reactions have been demonstrated by the presence of proteinuria and immune complex deposits in the kidneys of the Brown-Norway, Lewis, and PVG/C rat strains. However, susceptibility to the deposition and the subsequent lesions (glomerulonephritis) are often variable and dependent on the strain (Bigazzi, 1985). For example, despite the appearance of clinical signs and proteinuria, after two-months administration of mercuric chloride, detectable levels of circulating antinuclear autoantibodies can no longer be observed in the Brown-Norway strain (Bellon et al., 1982). By contrast, in PVG/C rats administered mercuric chloride, immune complex deposition and antinuclear autoantibodies are present for longer periods of time; however, proteinurea is not observed (Weening et al., 1978). Type IV hypersensitivity. There are several well-established preclinical models for assessing Type IV (delayed-type) hypersensitivity reactions following dermal exposure, but not for predicting this response after systemic exposure. The dermal exposure mode is the only currently required and widely performed immunotoxicity assay on devices. Type IV hypersensitivity responses are elicited by T lymphocytes and are controlled by accessory cells and

114

Safety Evaluation of Medical Devices

suppressor T cells. Macrophages are also involved in that they secrete several monokines, which results in proliferation and differentiation of T cells. Thus, there are numerous points along this intricate pathway in which drugs may modulate the final response. To achieve a Type IV response, an initial high-dose exposure or repeated lower-dose exposures are applied to the skin; the antigen is carried from the skin by Langerhans cells and presented to cells in the thymus to initiate T-cell proliferation and sensitization. Once sensitized, a second “challenge” dose will elicit an inflammatory response. Thus, before sensitivity can be assessed, each of the models used to evaluate dermal hypersensitivity requires as a minimum: ● ● ●

an initial induction exposure a latency period for expression a challenge exposure.

A preliminary test for acute irritancy is also required to ensure that the initial dose is sufficient to stimulate sensitization, and that the challenge dose is sufficient to ensure expression of the response without producing irritation, which would confound the response. To confirm suspected sensitization or determine a threshold dose, each assay may also include a second challenge dose one to two weeks after the first challenge, at the same or lower concentrations. To increase penetration of the test article, various methods of abrasion (e.g., tape stripping) and occlusive coverings may also be used. Assessing materials to determine if they can act as delayed contact dermal sensitizers in humans is different on a number of grounds from the other tests we have looked at so far and, indeed, from most of the other test systems presented later in this book. These differences all stem from how the immune system, which is the mechanistic basis for this set of adverse responses, functions. Bringing about this Coombs type IV hypersensitivity response (which is commonly called “sensitization” for short) requires more than a single exposure to the causative material, both in humans and in test animals. Unlike irritation responses, sensitization occurs in individuals in an extremely variable manner. A portion of the human population is considerably more liable to be sensitized while others are infrequently affected. And the response, once sensitization is achieved, becomes progressively more severe with each additional exposure. All three of these characteristics are due to the underlying mechanism for the response, and influenced the manner in which we conduct tests. These factors mean that in vivo test systems require multiple exposures of animals and tend to underpredict the potential for an adverse response in those individuals who are most susceptible to sensitization. But because the response to repeated exposures of even minimal amounts of material in these susceptible individuals can lead to such striking adverse responses, we must be concerned about them. A number of factors influence the potential for a chemical to be a sensitizer in humans and, in turn, also influence the performance of test systems. These are summarized in Table 9.12. Various test systems manipulate these in different ways. There are a number of references which explore and discuss the underlying immune system mechanisms and operation in greater detail. Particularly recommended is Gibson et al. (1983).

Table 9.12

Factors Influencing Delayed Type Sensitization Responses

Percutaneous absorption of agent. Genetic status of host. Immunological status of host. Host nutrition. Chemical and physical nature of potential sensitizing agent. Number, frequency and degree of exposures of immune system to potential antigen. 7. Concurrent immunological stimuli (such as adjuvants, inoculations, and infections). System can be “up modulated” by mild stimuli or overburdened by excessive stimulation. 8. Age, sex and pregnancy (by influencing factors 1, 3 and 4 above). 1. 2. 3. 4. 5. 6.

Objectives and General Features Given the considerations of mechanism, degree of concern about protecting people, and practicality, the desired characteristics of a sensitization test include the following. 1. Be reproducible. 2. Involve fairly low technical skills so that it may be performed as a general laboratory test. 3. Not involve the use of exotic animals or equipment. 4. Use relatively small amounts of test material. 5. Be capable of evaluating almost any material of interest. 6. Be sensitive enough to detect weak sensitizers (that is, those which would require extensive exposure to sensitize other than the most sensitive individuals). 7. Predict the relative potency of sensitizing agents accurately. Several of these desired characteristics are mutually contradictory; as with most other test systems, the methods for detecting dermal sensitization each incorporate a set of compromises. All the in vivo tests have some common features, however. The most striking is that they involve at least three (and frequently four) different phases—they are multiphasic. These phases are, in order, the irritation/toxicity screen, the induction phase, the challenge phase and (often) the rechallenge phase. Irritation/Toxicity Screen: All assays require knowledge of the dermal irritancy and systemic toxicity of the test material(s) to be used in the induction, challenge and rechallenge. These properties are defined in this pretest phase. Most tests desire (or will allow) mild irritation in the induction phase. Most tests desire (or will allow) mild irritation in the induction phase, but no systemic toxicity. Generally, a nonirritating concentration is required for the challenge and for any rechallenge, as having irritation present either confounds the results or precludes having a valid test. As will be discussed in the sections on the individual tests, even a carefully designed screen does not necessarily provide the desired guidance in selecting usable concentrations. During this phase, solvent systems are also selected. Induction Phase: This requires exposing the test animals to the test material several times over a period of days or weeks. A number of events must be accomplished during this phase if a sensitization response is to be elicited. The test material must penetrate through the epidermis and into the dermis. There, it must interact with dermis protein. The protein-test material complex must be perceived by the immune system as an allergen. Finally, the production of sensitized T cells must be accomplished. Some assays enhance the sensitivity of the induction phase by compromising the natural ability of the epidermis to act as a barrier. These enhancement

Chapter 9:

techniques include irritation of the induction site, intradermal injection, tape stripping and occlusive dressings. In contrast, events such as the development of a scab over the induction site may reduce percutaneous absorption. The attention of the immune system can be drawn to the induction site by the intradermal injection of oil-coated bacteria (Freund’s complete adjuvant, which serves as a mild immunological stimulant). Challenge Phase: This consists of exposing the animals to a concentration of the test material which would normally not be expected to cause a response (usually an erythema type response). The responses in the test animals and of the control animals are then scored or measured. Rechallenge Phase: This is a repeat of the challenge phase and can be a very valuable tool if used properly. Sensitized animals can be rechallenged with the same test material at the same concentration used in the challenge in order to assist in confirming sensitization. Sensitized animals can be rechallenged with different concentrations of the allergen to evaluate dose response relationships. Animals sensitized to an ingredient to evaluate can be challenged to a formulation containing the ingredient to evaluate the potential of the formulated product to elicit a sensitization response under adverse conditions. Conversely, animals which responded (sometimes unexpectedly) to a final formulation can be challenged with formulation without the suspected sensitizer or to the ingredient which is suspected to be the allergen. Cross reactivity can be evaluated. That is, the ability of one test material to elicit a sensitization response following exposure in the induction phase to a different test material. A well-designed rechallenge is important and should be considered at the same time that the sensitization evaluation is being designed, since the rechallenge must be run within 1 to 2 weeks after the primary challenge. Unless plans have been made for a possible rechallenge, one may have to reformulate a test material or obtain additional pure ingredient and perhaps run additional irritation/toxicity screens before the rechallenge can be run. The ability of the sensitized animals to respond at a rechallenge being run shortly after the challenge. In addition, some assays use sham-treated controls and these must be procured while the induction phase is in progress. One additional piece of information must be kept in mind when evaluating a rechallenge. The animal does not differentiate between an induction exposure and a challenge exposure. If one is using an assay which involves three induction exposures and one challenge exposure, then at the rechallenge, the animal has received four induction exposures. This “extra” induction may serve to strengthen a sensitization response. After the study is done, one must evaluate the data and decide how to translate it to human relevance. We will look at this problem towards the end of this chapter.

History Koch’s initial observation of tuberculin reactivity was made in humans and guinea pigs. Though the rabbit and guinea pig have both been considered the animals of choice for evaluating adverse skin reactions of chemicals, from the beginning guinea pigs have been the animals of choice for predictive tests. Though it is widely believed that this is due to a relatively higher degree of susceptibility to dermal sensitization, the preference was actually based on availability, ease of handling, and the fact that the albino animal has a clear pale skin which is easily denuded of hair and on which an erythema response is easy to distinguish.

Immunotoxicology

115

Landsteiner and Jacobs (1935) first proposed a formalized predictive test in guinea pigs. Later, he and Chase (1942) used low molecular weight chemicals to sensitize guinea pigs and developed the theory of complete antigen formations being due to hapten–protein interactions. The basis of modern predictive tests is the Draize tests, as established by Landsteiner and Draize et al. in 1944. It consists of 10 intradermal injections of the test compound into the skin of albino guinea pigs during the 3-week induction period and a single intracutaneous challenge application 14 days after the last induction injection. A standardized 0.1% test concentration is used for induction and challenge. This method was widely used and recommended until the end of the 1960s. Its disadvantage is that only strong allergens are detected, while well-known moderate allergens fail to sensitize the animals at all. Starting in 1964, however, a wide variety of new test designs started to be proposed. Buehler (1964, 1965) proposed what is now considered the first modern test (described in detail in this chapter), which used an occlusive patch to increase test sensitivity. The Buehler test is the primary example of the so-called “epidermal” methods, which have been criticized for giving false-negative results for moderate to weak sensitizers such as nickel. A new generation of tests was established by using Freund complete adjuvant (FCA) during the induction process to stimulate the immune system, independent of the type of hapten and independent of the method or application, that is, whether or not the substance is incorporated in the adjuvant mixture. It is claimed that this family of tests displays the same level of susceptibility to sensitization in guinea pigs as is normally observed in humans (Cronin and Agrup, 1970). The adjuvant tests include the guinea pig maximization test (Maurer et al., 1975, 1980), split adjuvant test (Maguire and Chase, 1967) and the epicutaneous maximization test (EMP, Guillot and Gonnet, 1985). Finally, during the last few years, a test system which uses albino mice instead of guinea pigs—the mouse ear swelling test (MEST, Gad et al., 1985a, b) and Local Lymph Node Assay (LLNA, Kimber et al., 1986)—have been adapted as alternatives by OECD, and the LLNA is specified as an appropriate test for pharmaceuticals (ICH) and medical devices (ISO). The LLNA, it should be noted, has the limitation that it only looks at the induction part of the paradigm. It does not include a close setting irritation probe first, and does not look at the elicitation phase at all. The benefits are that fewer animals are needed and the test takes less time. The cost is an increased false positive rate. This chain of development should be expected to continue and the overall quality and utility of tests should improve. Four tests will be presented and compared in this volume, as each has features and operating characteristics which both make them alliterative in particular circumstances and also cause them to be representative of other available tests. These are the Buehler, guinea pig maximization, split adjuvant and mouse ear swelling tests.

Modified Buehler Procedure This is a closed patch procedure for evaluating test substances for potential delayed contact dermal sensitization in guinea pigs. The procedure, based on that described by Buehler (1965), is practical for test substances that cannot be evaluated by the traditional intradermal injection procedure of Landsteiner and Jacobs or by the GPMT for skin sensitization testing. The closed patch procedure is performed when

116

Safety Evaluation of Medical Devices

Species: Guinea Pig Strain: Hartley Test group: 15 Animals Control group: 6 Animals Rechallenge Challenge (if necessary) Quarantine period −14

−7

Induction period 0

7

14

Rest period 21

Dermal applications of test material in suitable solvent/ carrier on shaved skin site for 6 hours under occlusive patch except for the first which is for 24 hours

Figure 9.3

28

35

42 (Day)

Closed patch application to naive skin site for 24 hours. Site is then scored at 24, 48 and 72 hours after removal of patch

Line chart of study design for modified Buehler test for delayed contact dermal sensitization in the guinea pig.

a test substance either is highly irritating to the skin by the intradermal injection route of exposure or cannot be dissolved or suspended in a form allowing injection. It is also the method of choice for some companies. This procedure, which is one version of the Buehler test, complies with the test standards set forth in the Toxic Substances Control Act (TSCA, 1979) and other regulatory test rules, and is presented diagramatically in Figure 9.3. There are other versions which also comply. Animals: 1. Young albino female guinea pigs, weighing between 300 and 400 g are used. 2. Although several proposed test rules suggest the use of male guinea pigs, the female sex is preferred because the aggressive social behavior of males may result in considerable skin damage that might interfere with the interpretation of challenge reactions. This concern occurs because animals are group housed (Marzulli and Maibach, 1996). 3. Animals that show poor growth or are ill in any way are not used, since illness markedly decreases the response. Animals with skin marked or scarred from fighting are avoided. The guinea pigs are observed for at least two weeks to detect any illness before starting a study. 4. The guinea pigs are identified by a cage card and marking pen or any other suitable method. There are no regulatory requirements, however, for the identification of individual animals. 5. The guinea pigs are randomly assigned to test and negative control group consisting of at least 15 and at least 6 animals each, respectively. If a pre-test group is necessary, as many animals as needed for that group are randomized also. Pre-Test Screen: 1. If practical, the dermal irritation threshold concentration should be established for the test substance prior to the first induction application. A concentration of the test substance that produces minimal or no irritation (erythema and/or edema formation) is determined. The highest concentration that produces no irritation is preferred for the dermal sensitization study challenge dose.

2. Those animals randomly assigned to the pre-test group are used. 3. Each animal is prepared by clipping a 1-inch square area of hair from the left upper flank using a small animal clipper with a size No. 40 blade. 4. The test substance is diluted, emulsified, or suspended in a suitable vehicle. Vehicles are selected on the basis of their solubilizing capacity for the test substance and on their ability to penetrate the skin. 5. Different concentrations of the test substance are tested on the pre-test group of guinea pigs; a few animals are used for each concentration tested. 6. A volume of 0.l5 mL is applied to a patch consisting of a cotton pad (2.5 × 2.5 cm) occluded with impermeable surgical tape, or placed in a Hilltop style occlusive “chamber.” 7. The patch is applied to the shaved left flank of a guinea pig. The patch is held firmly in place for 24 hours by wrapping the trunk of the animal with a 3-inch wide elastic bandage. A 2-inch wide strip of tape is used to line the center adhesive side of the bandage in order to prevent skin damage from the adhesive. 8. After 24 hours of exposure, the wrappings and patches are removed. 9. Observations of skin reactions (erythema and/or edema formation) are recorded 48 hours after application. 10. A judgment is made as to which concentration will be used for the dermal sensitization study, based on the dermal irritation data which has been collected. The highest concentration that produces minimal or no dermal irritation is selected. Induction Phase: 1. Test group and control group guinea pigs are weighed at the beginning of the study and weekly thereafter. 2. Test control group guinea pigs are clipped as described earlier in this procedure. 3. If the test substance is a liquid (solution, suspension, or emulsion), a volume of 0.5 mL of the highest concentration found to be nonirritating in a suitable vehicle (as determined in the pre-test portion of this procedure) is applied to a patch consisting of a cotton pad (1¢ × 1¢) occluded

Chapter 9:

Table 9.13

Evaluation of Skin Reactions (Draize, 1959)

Skin reaction Erythema and eschar formation: No erythema Very slight erythema (barely perceptible) Well-defined erythema Moderate to severe erythema Severe erythema (beef redness) to slight eschar formation (injuries in-depth) Necrosis (death of tissue) Eschar (sloughing) Edema formation: No edema Very slight edema (barely perceptible) Slight edema (edges of area well-defined by definite raising) Moderate edema (raised approximately 1 millimeter) Severe edema (raised more than 1 millimeter and extending beyond the area of exposure)

Value 0 1 2 3 4

2.

+N +E

3.

0 1 2 3 4

4. 5. 6.

4.

5. 6.

7.

8. 9.

with impermeable surgical tape. If the test substance is a solid or semi-solid 0.5 g is applied.a If the test substance is a fabric, a 1-inch square is moistened with 0.5 mL of physiological saline before application. The first induction patch is applied to the clipped left flank of each test group guinea pig. The patch is held firmly in place for 24 hours by wrapping the trunk of each animal with a 3-inch wide elastic bandage. A 2-inch wide strip of tape is used to line the center adhesive side of the bandage in order to prevent skin damage from the adhesive. A 2-inch length of athletic adhesive tape is placed over the bandage wrap as a precautionary measure to prevent unraveling. After 24 hours of exposure, the wrappings and patches are removed and disposed of in a plastic bag. Each dermal reaction, if any, is scored on the basis of the designated values for erythema and edema formation presented in Table 9.13. Observations are made 48 hours after initiation of the first induction application. Resulting dermal irritation scores are recorded. After the initial induction application, subsequent induction applications (2–10) are made on alternate days (three times weekly) until a total of 10 treatments is administered. Each of these patches is removed after 6 hours of exposure. It should be noted that some use a modification which calls for one application per week for 3 weeks. Observations are made 24 and 48 hours after initiation of each subsequent induction application. Dermal scores of the remaining nine induction applications are recorded. Clipping the hair from the left flank induction sites of test group animals and corresponding sites on negative control group animals is performed just prior to each subsequent induction application. Only the test group guinea pigs receive the induction applications.

Challenge Phase: 1. Fourteen days after the tenth induction application, all 10 test group, and 3 of 5 control group, guinea pigs are a

When the test substance is in flake, granule, powder or other particulate form, the weight of the test substance that has a volume of 0.5 mL (after compacting as much as possible without crushing or a altering the individual particles, such as by tapping the measuring container) is used whenever this volume weighs less than 0.5 g.

Immunotoxicology

117

prepared for challenge application by clipping a 1-inch square of hair from the right side, the side opposite that which was clipped during the induction phase. A challenge dose, using freshly prepared test substance (solution, suspension, emulsion, semi-solid, solid or fabric), is applied topically to the right side (which had remained untreated during the induction application) of test group animals. The left side, which had previ ously received induction applications, is not challenge dosed. The concentration of the challenge dose is the same as that used for the first induction application. (It must be a concentration that does not produce dermal irritation after one 24 hr application.) Each of three negative control group guinea pigs is challenge dosed on the right flank at approximately the same time that the test group guinea pigs are challenge dosed. All patches are held in contact with the skin for 24 hours before removal. The skin sites are evaluated using the scoring system for erythema and edema formation presented in Table 9.13. Observations are made 48, 72, and 96 hours after initiation of the challenge application. Skin reactions are recorded.

Rechallenge Phase: 1. If the test substance is judged a nonsensitizing agent after the first challenge application, or causes dermal sensitization in only a few animals or causes dermal reactions that are weak or questionable, then a second and final challenge application will be performed on each test animal 7 days after the initiation of the first challenge dose. 2. Controls from the first challenge application are not rechallenged because they have been exposed to the test substances and are no longer true negative controls. The three remaining naive control group animals (not used for the first challenge) are challenged for comparison to the test group animals. 3. The procedure used for the first challenge application will be used for the second challenge application (including reclipping, patching method, and duration of exposure). Either the same concentration or a new concentration (higher or lower) of test substances may be used, depending on the results of the first challenge. Observations are made 48, 72, and 96 hours after initiation of the rechallenge application and skin reactions are recorded. 4. When a rechallenge application is performed, the data from both challenges are compared. If neither challenge produces a positive dermal reaction, the classification of the test substance is based on both challenge applications. If one challenge application (whether it is the first or second) produces a greater number of positive dermal reactions than the other, the classification of the test substance is based on the challenge with the most positive responses. 5. Two or more unequivocally positive responses in a group of 15 animals should be considered significant. A negative, equivocal, or single response probably assures that a substance is not a strong sensitizer, although this is best confirmed by further testing with human subjects (NAS, 1977). Interpretation of Results: 1. Judgment concerning the presence or absence of sensitization is made for each animal. The judgment is made by comparing the test animal’s challenge responses to its first

118

Safety Evaluation of Medical Devices

Stage Day

TEST GROUP (15)

TEST GROUP (15)

Figure 9.4

INDUCTION 0

CHALLENGE 7

A. 0.1 ml Substance ID B. 0.1 ml FCA ID C. 0.1 ml Substance + FCA ID

RECHALLENGE

21

28

Closed patch-48 H application of Substance

Closed patch-24 H Vehicle Substance

Closed patch-24 H Vehicle

Closed patch-48 H application of Vehicle

Closed patch-24 H Substance Vehicle

Closed patch-24 H Substance

A B C

A. 0.1 ml Vehicle ID B. 0.1 ml FCA ID C. 0.1 ml Vehicle + FCA ID A B C

Line chart of study design for guinea pig maximization test (GPMT) for predicting delayed dermal sensitization.

induction treatment response, as well as to those challenge responses of negative control animals. 2. Challenge reactions to the test substance that are stronger than challenge reactions to negative controls or to those seen after the initial induction application should be suspected as a result of sensitization (NAS, 1977). A reaction that occurs at 48 hours, but resolves by 72 hours or 96 hours, should be considered a positive response as long as it is stronger than that which is displayed by controls at the same time interval.

laboratories keep a “pool” of guinea pigs on-hand (especially as they are the most expensive of the common laboratory species), the usual case is that 8–10 weeks is the minimum time required to get an answer from this test. 3. The test uses a relatively large amount of test material. In the normal acute “battery,” the guinea pig test systems use more material than any other test systems unless an acute inhalation study is included. With 10 induction applications, this is particularly true for the Buehler style test.

Guinea Pig Maximization Test Strengths and Weaknesses: There are a number of both advantages and disadvantages to the Buehler methodology, which has been in use for 20 years. The relative importance and merits of each depend on the intended use of the material. The four advantages are: 1. Virtually no false positives (in fact, in the experience of the author when the pretest is properly conducted, there are no false positives), compared to human experience, are generated by this test. 2. The techniques involved are easy to learn and very reproducible. 3. The Buehler style test does not overpredict the potency of sensitizers. That is, materials which are identified as sensitizers are truly classified as very strong, weak, or in-between—not all (or nearly all) as very strong. 4. There is a large data base in existence for the Buehler style test. Unfortunately, the vast majority is not in the published literature. Likewise, there are three disadvantages associated with the Buehler style test. 1. The test gives a high rate of false negatives for weak sensitizers and a detectable rate of false negatives for moderate sensitizers. That is, the method is somewhat insensitive—particularly if techniques for occlusive wrapping are inadequate. 2. The test takes a long time to complete. If animals are on-hand when started, the test is 5 to 6 weeks long. As few

The Guinea Pig Maximization Test (GPMT) was developed by Magnusson and Kligman (1969, 1970; Magnusson, 1975) and is considered a highly sensitive procedure for evaluating test substances for potential dermal sensitization. The procedure presented here is illustrated diagrammatically in Figures 9.4 and 9.5, and is one common version of the test. Animals: 1. Young adult female guinea pigs, weighing between 250–350 g at the initiation of the study, are used. 2. Although several proposed test rules suggest the use of male guinea pigs, the female sex is preferred because the aggressive social behavior of males may result in considerable skin damage that might interfere with the interpretation of challenge reactions. 3. Animals that show poor growth or are ill in any way are not used, since illness markedly decreases the response. Animals with skin marked or scarred from fighting are avoided. The guinea pigs are observed for at least 2 weeks to detect any illness before starting the study. 4. The guinea pigs are randomly assigned to two groups: (1) a test group consisting of 15 animals; and (2) a control group consisting of 6 animals. If a pre-test group is necessary, as many animals as needed for that group are randomized also. 5. Test and control group guinea pigs are weighed one week prior to dosing (day 7), on the day of dosing (day 0), and weekly thereafter.

Chapter 9:

Immunotoxicology

119

Species: Guinea Pig Strain: Hartley Test group: 15 Animals Control group: 6 Animals Challenge Quarantine period −14

−7

2X ID Injections of a) Substance in vehicle b) FCA c) Substance in FCA

Figure 9.5

Induction period 0

Rest period

7

14

Closed patch application to ID injection skin site for 48 hours

Rechallenge (if necessary) 21

28 (Day)

Closed patch application to naive skin site for 24 hours

Illustrative figures for injection and patching of animals in GPMT.

Pretest: 1. Several animals are used to pretest the test substance and vehicles to determine the topical dermal irritation threshold concentration. 2. These animals are shaved on the left flank, to which is applied a 2 × 2 cm filter paper patch which contains 0.1 mL of test concentration. 3. The trunks of the animals are wrapped for 24 hours with a 3-inch wide elastic bandage to hold the patch in contact with the skin. 4. Wrappings are removed after the 24 hr exposure and, based on skin reactions at 48 hr, a concentration of the test substance to be used on test is determined. Dermal irritation values are recorded for future reference. 5. In addition, several guinea pigs are utilized to determine a concentration (generally, between 1–5%) of test substance in vehicle and in FCA emulsion that can be injected id without eliciting a strong local or systemic toxic reaction. 6. The hair is clipped in an area of approximately 4 × 6 cm from the upper shoulder region of these animals. 7. Several concentrations of test substances (ranging between 1–5%) can be injected in the same animal to compare local dermal reactions produced by the different concentrations. 8. However, if systemic toxicity is suspected, then each concentration should be tested in separate animals to determine local and systemic effects. 9. The dermal reactions (erythema, edema, and diameter) are recorded 24 hours after the id injections. Induction Stage 1 (Day 0): 1. The hair in an area of 4 × 6 cm is clipped from the shoulder region of each test and control group guinea pig on day 0. 2. Three pairs of intradermal (id) injections are made with a glass 1 ml tuberculin syringe with a 26-gauge needle, each pair flanking the dorsal midline. 3. The three pairs of id injections for test group animals are as follows: (a) 0.1 mL test substance in appropriate vehicle (b) 0.1 mL Freund’s Complete Adjuvant (FCA) Emulsion alone (c) 0.1 mL test substance in FCA Emulsion.

4. The three pairs of id injection for control group animals are as follows: (a) 0.1 mL vehicle alone (b) 0.1 FCA Emulsion alone (c) 0.1 mL vehicle in FCA Emulsion. 5. Injections (a) and (b) in the above two steps are given close to each other and nearest the head; injection (c) is given most posteriorly. 6. The date, time, and initials of those individuals performing the id injections are recorded. Immediately before injection, an emulsion is prepared by blending commercial FCA with an equal volume of house distilled water or other solvent as appropriate. (a) Water-soluble test materials are dissolved in the water phase prior to emulsification. (b) Oil-soluble or water-insoluble materials are dissolved or suspended in FCA prior to adding water. (c) Paraffin oil, peanut oil, or propylene glycol can be used for dissolving or suspending water-insoluble materials. (d) A homogenizer is used to emulsify the FCA alone and the test substance in either FCA or vehicle prior to the id injections. (e) The concentration of the test substance for id injections is adjusted to the highest level that can be well tolerated locally and generally. 8. The adjuvant injection infiltration sometimes causes ulceration, especially when it is superficial, which lasts several weeks. These lesions are undesirable but do not invalidate the test results except for lowering the threshold level for skin irritation. Induction Stage 2 (Day 7): 1. Test Substance Preparation (a) The concentration of the test substance is adjusted to the highest level that can be well tolerated. (b) If the test substance is an irritant, a concentration is chosen that causes a weak to moderate inflammation (as determined by the pretest). (c) Solids are micronized or reduced to a fine powder and then suspended in a vehicle, such as petrolatum or propylene glycol.

120

2. 3. 4. 5. 6.

Safety Evaluation of Medical Devices

(d) Water-and-oil soluble test substances are dissolved in an appropriate vehicle. (e) The concentration of the test substance for id injections is adjusted to the highest level that can be well tolerated locally and generally. The same area over the shoulder region that received id injections on day 0 is again shaved on both test and control guinea pigs. A volume of 0.3 mL of a mildly irritating concentration (if possible) of the test substance (determined by the pretest) is spread over a 1 × 2 inch filter of each test group animal. The control group animals are exposed to 0.3 mL of 100% vehicle using the same procedure. The date, time, and initials of those individuals performing the second induction are recorded. The dressings of both groups are left in place for 48 hours before removal.

Challenge Stage (Day 21): 1. An area of hair (1.5 × 1.5 in.) on both flanks of the guinea pigs (15 test and 3 controls) is shaved. 2. A 1 × 1 inch patch with a nonirritating concentration of test substance in vehicle (as determined by the pretest) is applied to the left flank and a 1 × 1 inch patch with 100% vehicle is applied to the right flank. 3. The torso of each guinea pig is wrapped in an elastic bandage to secure the patches for 24 hours. 4. The date, time, and initials of those individuals performing the challenge dose are recorded. 5. The patches are removed 24 hours after application. Rechallenge (Day 28): 1. If the first challenge application of test substance does not cause dermal sensitization, causes dermal sensitization in only a few animals, or causes dermal reactions that are weak or questionable, then a second challenge application of test substance to the 15 test group guinea pigs will be conducted on day 28 (one week after the first challenge). The three remaining naive control group animals (not used for the first challenge) are challenged for comparison to the test group animals. 2. The three negative control group animals used on day 21 will not be rechallenged. These animals will be terminated because they were exposed to the test substance during the first challenge and are no longer negative controls. 3. A 1 × 1 inch patch with a nonirritating concentration of test substance in vehicle is applied to the right flank of test and control group animals. The left flanks are not dosed. 4. The date, time, and initials of those individuals performing the rechallenge dose are recorded. 5. Steps 3 and 5 are followed as for Challenge State (Day 21). Observations—Challenge and/or Rechallenge Readings: 1. Twenty-one hours after removing the patch, the challenge area on each flank is cleaned and clipped, if necessary. 2. Twenty-four hours after removing the patch, the first reading of dermal reactions is taken. 3. The dermal reactions are scored on a 4-point scale (as below): 0—no reaction 1—scattered mild redness 2—moderate and diffuse redness 3—intense redness and swelling 4. Forty-eight hours after removing the patch, the second reading is taken and the scores are recorded.

Table 9.14 Sensitization Severity Grading Based on Incidence of Positive Responses (Kligman, 1966) Sensitization rate % 0–8 9–28 29–64 65–80 81–100

Grade I II III IV V

Classification Weak Mild Moderate Strong Extreme

Interpretation of Results: 1. Both the intensity and duration of the test responses to the test substance and the vehicle are evaluated. 2. The important statistic in the GPMT is the frequency of sensitization and not the intensity of challenge responses. A value of 1 is considered just as positive as a value of 3 (as long as the values for controls are zero). 3. The test agent is a sensitizer if the challenge reactions in the test group clearly outweigh those in the control group. A reaction that occurs at 24 hours, but resolves by 48 hours after removal of patches, should be considered a positive response, as long as it is stronger than that which is displayed by controls. The sensitization rate (% of positive responders) is based on the greatest number of animals showing a positive response, whether it is from the 24-hour data or the 48-hour data after removal of patches. 4. When a second challenge application is performed, the data from both challenges are compared. If neither challenge produces a positive dermal reaction, the classification of the test substance is based on both challenge applications. If one challenge application (whether it is the first or second) produces a greater number of positive dermal reactions than the other, the classification of the test substance is based on the challenge with the most positive responses. 5. Under the classification scheme of Kligman (1966, shown in Table 9.14), the test substance is assigned to one of five classes, according to the percentage of animals sensitized, ranging from a weak grade I to an extreme grade V. The advantages and disadvantages of the GPMT can be summarized as follows. First, the advantages: 1. The test system is sensitive and effectively detects weak sensitizers. If has a low false negative rate. 2. If properly conducted, there are no false positives—that is, materials which are identified as potential sensitizers will act as such at some incidence level in humans. 3. There is a large database available on the evaluation of compounds in this test system, and many people are familiar with the test system. The disadvantages, meanwhile, are: 1. The test system is sensitive; it overpredicts potency for many sensitizers. There is no real differentiation between weak, moderate, and strong sensitizers; virtually all positive test results identify a material as strong. 2. The techniques involved (particularly the intradermal injections) are not easy. Some regulatory officials have estimated that as many as 35% of the laboratories which try cannot master the system to get it to work reproducibly. 3. The test, though not as long as the Buehler, still takes a minimum of four weeks to produce an answer. 4. The test uses a significant amount of test material.

Chapter 9:

Immunotoxicology

121

Challenge stage Quarantine period

−7

Induction stage

0

1

3

10 11 12

5

Topical Topical application to application 24h 48h abdominal skin prepared to ears Ear thickness by FCA ID injection measurement on day 0 and tape stripping on all 4 days Animals receive diet supplemented with 250 IU/g of feed of vitamin A acetate

Figure 9.6

Line chart of study design for optimized mouse ear swelling test (MEST).

5. One cannot evaluate fibers or other materials which cannot be injected (such as solids which cannot be finely ground and/or suspended or which are highly irritating or toxic by the IV. route). 6. The irritation pretest is critical. Failure to detect irritation in this small group of animals does not guarantee against irritation in test animals at challenge.

Mouse Ear Swelling Test (MEST) Several of the disadvantages associated with the previous three test systems, both stated and unstated, are reflections of limitations of the guinea pig as a model and the methodology of evaluating response in terms of observing and subjectively “grading” skin erythema. Since Crowle (1959a, b) formally proved that passive transfer of delayed-type contact hypersensitivity exists in the mouse in 1959, research immunologists have generated

Day

0

1,3 and 5

1) Fur of abdomen is clipped 2) ID Injection of FCA (Freund’s complete adjuvant) 3) Abdominal skin is tape stripped 4) Topical application of substance of vehicle 5) Abdominal skin site is dried rapidly (electric dryer) Test group

1) Abdominal skin is tape stripped 2) Topical application of substance of vehicle 3) Abdominal skin site is dried rapidly

Control group 100 µL of test substance in vehicle

Figure 9.7

a wealth of information in attempts to understand the delayed-type hypersensitivity (DTH) response in this species (Asherson and Ptak, 1968). In particular, they have demonstrated that thymusderived cells are necessary for inducing a DTH response (DeSousa and Parrott, 1969). Also, the mouse has been used to investigate immunosuppressive properties of certain drugs, such as fluorinated steroids and corticosteroids. All of these have led to the development of a formalized test procedure, the MEST. The MEST is a procedure based on that which is described by Gad et al. (1985a, b, 1986, 1987) for evaluating test substances for their potential to cause dermal sensitization in mice. This procedure evaluates contact sensitization by quantitatively measuring mouse ear thickness. This method is shown diagrammatically in Figures 9.6–9.8.

Illustrative diagram of MEST induction stage procedures.

100 µL of vehicle

122

Safety Evaluation of Medical Devices

Day

10

11 & 12

1) Topical application of test substance to one ear 2) Topical application of vehicle to contralateral ear 3) Both ears are dried rapidly

40 µL of test substance in vehicle

40 µL of vehicle

Test and control animals

Figure 9.8

Ear thickness measurement of test and control ears is made with micrometer 24 and 48 hours after exposure

Areas of measurement

Test and control animals

Illustrative diagram of MEST challenge stage procedures.

Animals: 1. CF-1 or Balb/C female mice, 6–8 weeks old, are used. The mice are observed for at least one week to detect any signs of illness before starting a study. Mice that show poor growth or signs of illness are excluded from use on a test. 2. Any mouse displaying redness on either ear prior to the start of a test should be replaced. 3. Mice, which have been randomly placed in cages upon arrival, are assigned to groups (a maximum of five per cage) by labeling cage cards. For each test substance investigated, a pretest group of at least 8 mice, a test group of at least 15 mice, and a control group of at least 10 mice are utilized. For two weeks prior to initiation of testing (starting on arrival), animals are given diet supplemented with vitamin A (Thorne, 1992). 4. Because animals are not individually marked, they will always be handled one at a time when each phase of this procedure is performed. The following procedure is conducted to prevent mixing animals during each phase (e.g., shaving, id injections, tape-stripping, and dosing). All mice are removed from their original cage and placed in an empty cage for holding. One mouse is removed from the holding cage at a time, the phase activity is performed, and then the mouse is returned to its original cage. This step will be repeated for each of the remaining mice in the holding cage. Pretest: 1. A dermal (abdomen and ear) irritation and toxicity probe study is conducted one week prior to the actual MEST in order to establish the maximum concentration of test substance that produces minimal irritation to the abdomen (belly) region after a single topical application on each of four days (if the substance does have potential to irritate skin) and to establish a concentration of test substance that is nonirritating to the ear after a single topical application. Also, dose levels of the test substance that produce systemic toxicity can be established during the pretest.

2. The test substance is diluted, emulsified, or suspended in a suitable vehicle. A vehicle (such as acetone, 70% ethanol, 25% ethanol, or methyl ethyl ketone) is selected which should be able to solubilize the test substance and be volatile. 3. Two mice from the pretest group are used to test each concentration of test substance. As many as four concentrations can be evaluated. The same mice used for belly irritation are also used for ear irritation testing. 4. On day 0, the first day of the pretest, each animal is prepared by clipping the hair from the belly region using a small animal clipper with a size No. 40 blade. 5. After clipping the belly, the outer layers of epidermis (stratum corneum) of each mouse are removed from the shaved belly region with a tacky transparent tape (1-inch wide). This procedure is referred to as “tape-stripping.” On day 0, the belly skin of each mouse is tape-stripped until the application region appears shiny. While an assistant supports the dorsal portion of the mouse, the tape is pressed firmly over the clipped belly region and quickly removed; this procedure is repeated as many times as needed. 6. After tape-stripping the belly, a volume of 100 µL of test substance is applied to the belly region using a microliter pipette. At the same time, test substance is applied to the ventral surface (10 µL) and dorsal surface (10 µL) of the left ear of the mouse using a microliter pipette. 7. On day 1, 24 hours after dosing the ears, the thickness of all probe animal ears is measured using an Oditest Model D-1000 thickness gauge. (a) Ether is used to anesthetize the mice in a fume hood while the ears are measured. (b) When a mouse reaches the “surgical anesthesia” stage, it is removed from the jar and gently placed on the countertop of the fume hood, which is prepared with a protective lining. (c) While supporting the mouse with one hand, the other hand is used to press the finger lever on the Oditest gage in order to open the flat measurement contacts. One ear of the mouse is then inserted between the

Chapter 9:

8.

9. 10. 11.

contacts until it is positioned with approximately 1–2 mm of the outer edge of the ear showing. After positioning the ear, the finger lever is released to allow the contacts to clamp onto the ear. The measurement is read from the gage after the indicator needle is stabilized. If desired, one or two more measurements can be rapidly made to be certain of a reproducible reading. (d) Once a reading is obtained, the other (contralateral) ear is measured in the same manner. The animal’s body is turned over in order to position the other ear for measurement. (e) Measurements are recorded. On subsequent days 1, 2, and 3, the belly region is first tape-stripped five times and then a volume of 100 µL of test substance is applied topically to the belly region using a microliter pipette. On day 4, 24 hours after the last topical application, the belly skin of all animals is observed for dermal irritation. A description of the results is recorded. If any signs of systemic toxicity are observed on any of the pretest days, then they should be noted. Based on the results of the pretest data, a judgment is made as to which concentration will be used for topical induction applications to the belly and for topical challenge application to the ear. A minimal or mildly irritating concentration is preferable for induction, and the highest nonirritating concentration possible is used for challenge application.

Induction Stage: 1. Day 0: (a) The belly of each test and control group mouse is clipped free of hair. (b) Immediately after clipping, two id injections of FCA emulsion are made at separate sites in the skin of the shaved belly (each site flanks the ventral mid-line). Approximately 20 µL of FCA emulsion is injected with a glass tuberculin syringe with a 30-gauge needle attached. Injections are performed in test and control mice. (c) Following the id injections, the belly skin of test and control group animals is tape-stripped until the site gives a shiny appearance. (d) After tape-stripping the belly, a volume of 100 µL of test substance (at a concentration determined by pretest) is topically applied to the belly skin of test group animals with a microliter pipette. Control animals receive a dose of 100 µL of vehicle. 2. Days 1, 2, and 3: (a) The skin of the belly of test and control group animals is tape-stripped five times. (b) After tape-stripping, a volume of 100 µL of test substance is topically applied to the belly skin of test group animals and a volume of 100 µL of vehicle/solvent is topically applied to control group animals. Challenge Stage: 1. Day 10: Each test group mouse and each of 5 control group mice is dosed with 10 µL of a concentration of test substance (determined by the pretest data) on the ventral side of the left ear and 10 µL on the dorsal side of the left ear.

Immunotoxicology

123

The contralateral right ear is dosed with 10 µL of 100% vehicle on the ventral side and 10 µL on the dorsal side. 2. Day 11: Ear thickness measurements are made 24 hours after challenge dosing. The procedure described in section 7, a, b, c, d of this SOP is used. 3. Day 12: Each thickness measurement is made again 48 hours after challenge dosing. Rechallenge: 1. If the test substance is judged a nonsensitizing agent after the first challenge application, causes dermal sensitization in only a few animals, or causes ear swelling that is weak or questionable, then a second and final challenge application will be performed on each test animal on day 17. 2. The five control group mice from the first challenge are not rechallenged because they have been exposed to the test substance and are no longer true negative controls. The five remaining naive control group animals (not used for the first challenge) are challenged for comparison to the test group animals. 3. The procedure used for the first challenge application will be used for the rechallenge application. Either the same concentration or a new concentration (higher or lower) of test substance may be used, depending on the results of the first challenge. 4. Measurement of both ears is performed on days 18 and 19 (24 and 48 hours after rechallenging, respectively). Each thickness measurement is recorded. Interpretation of Results: 1. Judgment concerning the presence or absence of sensitization is made for each animal. The judgment is based on the percent difference (%䉮) between test and control ears. A “positive” sensitization response is considered to have occurred if the test ear of an animal is at least 20% thicker than the control ear. 2. The percentage of animals in a test group that is considered “positive” is then calculated and recorded as percent responders. 3. The negative control group ear thickness measurements are used to identify any possible dermal irritation reactions, which would be interpreted as false positive dermal sensitization responses. 4. In addition, % ear swelling is calculated for the test group. The left (A) and right (B) ear thickness measurements are added. Percent ear swelling equals the sum of A (test ear thicknesses) divided by the sum of B (control ear thicknesses), multiplied by 100. % Ear Swelling = A/B × 100 5. When a second challenge application is performed, the data from both challenges are compared. If neither challenge procedure produces a positive sensitization reaction, the classification of the test substance is based on both challenge applications. If one challenge application (whether it is the first or second) produces a greater number of positive dermal reactions than the other, the classification of the test substance is based on the challenge with the most positive responses.

124

Safety Evaluation of Medical Devices

−7

−1

1

2

3

4

5

6

Study day

Animal receipt

Figure 9.9

Assign Treat 25 µL/day animals dorsal surface of each ear

Inject 3H-thymidine necropsy animals Five hours later, remove and process auricular lymph nodes

The mouse local lymph node assay.

6. Two or more unequivocally positive responses in a group of 15 animals should be considered significant. A negative, equivocal, or single response probably assures that a substance is not a strong sensitizer, although this is best confirmed by further testing with human subjects (NAS, 1977). Strengths and Weaknesses: The MEST offers distinct advantages compared to the guinea pig dermal sensitization procedures: 1. 2. 3. 4. 5. 6. 7.

The mouse is markedly less expensive. Less vivarium space is required. The duration of the test is shorter. Less test substance is utilized. Overall cost of the test is significantly less. The test is objective, rather than subjective. Materials which stain the skin may be easily evaluated. Several of the materials evaluated were colored and very difficult to evaluate by existing methods. 8. The test has a low false negative rate and no false positive rate, if properly performed. 9. The test seems to do a more accurate job of predicting relative hazards to humans. Disadvantages include the following: 1. The data base, though now not small, is not as extensive as that for GPMT or Buehler tests. 2. Fewer people have experience with the test system. 3 Regulatory status of data from the test system is unclear at present.

LOCAL LYMPH NODE ASSAY (LLNA) This method has developed out of the work of Ian Kimber and associates (Kimber et al., 1986; Kimber and Dearman, 1994; Kimber and Weisenberger, 1989). It has the advantage over the other methods discussed in this chapter in that it provides an objective and quantifiable endpoint. The method is based on the fact that dermal sensitization requires the elicitation of an immune response. This immune response requires proliferation of a lymphocyte subpopulation. The local lymph node assay (LLNA) relies on the detection of increased DNA synthesis via titriated thymidine incorporation. Sensitization is measured as a function of lymph node cell proliferative responses induced in a draining lymph node following repeated topical exposure of the test animal to the test article. Unlike the other tests discussed in this

chapter, this assay looks only at induction because there is no challenge phase. The typical test (illustrated in Figure 9.9) is performed using mice—normally female CBA mice 6–10 weeks of age. Female BALB/c and ICR mice have also been used. After animal receipt, they are typically acclimated to standard laboratory husbandry conditions for 7–10 days. The usual protocol will consist of at least two groups (vehicle control and test article treated) of five mice each. They are treated on the dorsal surface of both ears with 25 µL (on each ear) of test article solution for three consecutive days. After 24–48 hours of the last test article exposure, the animals are given a bolus (0.25 mL) dose of [3H]thymidine (20 µCi with a specific activity of 5.0–7.0 Ci/mmol) in phosphate buffered saline via a tail vein. Five hours after the injection, the animals are euthanized by CO2 asphyxiation and the auricular lymph nodes removed. After removal, the lymph nodes can either be pooled by group or processed individually. Single cell suspensions are prepared by gentle mechanical disaggregation through a nylon (100 µm) mesh. Cells are washed twice by centrifugation in an excess of PBS. After the final supernatant wash is removed, the cells are precipitated with cold 5% trichloroacetic acid (TCA) and kept at 4ºC for 12–18 hours. The precipitate is then pelleted by centrifugation and resuspended in 1 mL 5% TCA, and the amount of radioactivity is determined by liquid scintillation counting, using established techniques for tritium. The data are reduced to the stimulation index (SI): SI =

H (dpm treated group) H (dpm control group)

An SI of 3 or greater is considered a positive response, i.e., the data support the hypothesis that the test material is a sensistizer. The test article concentration is normally the highest nonirritating concentration. Several concentrations could be tested at the same time should one wish to establish a dose– response curve for induction. The test is easiest to perform if the vehicle is a standard non-irritating organic, such as acetone, ethanol, or dimethylformamide, or a solvent-olive old blend. Until a laboratory develops its own historical control base, it is also preferable to include a positive control group. Either 0.25% dinitricholorobenzene or 0.05% oxazalone are recommended for positive controls. If the vehicle for the positive control is different than the vehicle for the test material, then two vehicle control groups may be necessary.

Chapter 9:

Species: Strain: Test Group: Control Group:

Immunotoxicology

125

Mouse CBA/Ca 6 Animals 6 Vehicle

Induction period labeling

Quarantine period

(Days) −7

−1

0

1

2

3

Treat dorsal surface of each ear with 25 µL on three consecutive days

4

5

6

Inject 3H-thymidine necropsy animals 5 hr later, remove and process auricular lymph nodes.

Modification using flow cytometry instead of radiolabeling is preferable.

Figure 9.10

Mouse local lymph node assay (LLNA) (ICVAM protocol).

This method has been extensively validated in two international laboratory exercises (Basketter et al., 1991; Loveless et al., 1996). In the earlier work (Basketter et al., 1991), there was good correlation between the results obtained with guinea pig tests and those obtained with the LLNA. In a recent report, for example, five laboratories correctly identified dinitrochlorobenzene and oxazalone as sensitizers and the fact that p-aminobenzoic acid was not (Loveless et al., 1996). Arts and colleagues (1996) demonstrated that rats could be used as well as mice. Interestingly, they validated their assay (for both rats and mice) using BrDU uptake and immunohistochemical staining (rather than [3H] thymidine) to quantitated lymph node cell proliferation. This method is relatively quick and inexpensive because it uses relatively few mice (which are much less expensive than guinea pigs) and takes considerably less time than traditional guinea pig assays. It has an advantage over other methods in that it does not depend on a somewhat subjective scoring system and produces an objective and quantifiable endpoint. It does require a radiochemistry laboratory. Unless one already has an appropriately equipped laboratory used for other purposes (most likely metabolism studies), setting one up for the sole purpose of running the LLNA does not make economic sense. The standard version of the test has been adopted by OECD, ISO, ICH, and ICVAM, but also has been shown to have a modest false positive rate (misidentifying strong irritants as sensitizers). There is a modified version which (using flow cytometry and/or measurement of cytoline levels) is believed to solve this false positive rate problem.

Test System Manipulation (for all in vivo test systems) Increasing percutaneous absorption will increase test sensitivity. Factors which will increase absorption (and techniques for achieving them) include the following: 1. Increase surface area of solids. 2. Hydrate region of skin exposed to chemical. This can be done by wetting solids and using very occlusive wrapping of application. 3. Irritate application site.

4. Abrade application site. 5. Injection of test material (if possible). 6. Proper selection of solvent or suspending system. (See Christensen et al. (1984) for a discussion of the effect of vehicle in the case of even a strong sensitizer.) 7. Remove part or all of the “barrier layer” (stratum corneum) by tape stripping the application site. 8. Increase the number of induction applications. Though it is not a factor which increases percutaneous absorption, mildly stimulating the immune system of test animals (by such means as injecting FCA or some other adjuvant alone or FCA blended with the test material) also increases responsiveness to the test system. Also, it is generally believed that using the highest possible test material concentrations (mildly irritating for induction, just below irritating for challenge) will guarantee the greatest possible sensitization response, and will therefore also serve to universally increase sensitivity. There are reports, however (Gad et al., 1985a for croton oil and Thorne et al., 1986 for isocyanates) that this is not true for all compounds and that a multiple dose (i.e., two or more concentrations) study design would increase sensitivity. Such designs, however, would also significantly increase cost. Concurrent or frequent positive and negative controls are essential to guard against test system failure. Any of these test systems should show 0.05% dinitrochlorobenzene (DNCB) in 70% ethanol to be a strong sensitizer. In Vitro methods: There are actually several approaches available to in vitro evaluation of materials for sensitizing potential. These use cultured cells from various sources and, as endpoints, look at either biochemical factors (such as production of MIF-migration inhibition factor) or cellular events (such as cell migration or cell “transformation”). Milner (1970) reported that lymphocytes from guinea pigs sensitized to dinitrofluorobenzene (DNFB) would transform in culture, as measured by the incorporation of tritiated thymidine, when exposed to epidermal proteins conjugated with DNFB. This work was later extended to guinea pigs sensitized to p-phenylenediamine. He later (1971) reported that his method was capable of detecting

126

Safety Evaluation of Medical Devices

Table 9.15 Requested Reference Compounds for Skin Sensitization Studies (U.S. Consumer Product Safety Commission) Tribromophylophosphate Ditallow dimethyl ammonium methyl sulfate Hydroxylamine sulfate Ethyl amino benzoate Todochlorohydroxy quinoline (clioquinol, chinoform) Nickel sulfate Monomethyl methacrylate Mercaptobenzothiazole

Formalin Turpentine Potassium dichromate Penicillin G p-Phenylenediamine Epoxy systems (ethylenediamine, diethylenetriamine, and diglycidyl ethers) Toleune 2,4-diisocyanate Oil of Bergamot

allergic contact hypersensitivity to DNFB in humans when he used human lymphocytes from sensitized donors and human epidermal extracts conjugated with DNFB. Miller and Levis (1973) reported the in vitro detection of allergic contact hypersensitivity to DNCB conjugated to leukocyte and erythrocyte cellular membranes. This indicated that reaction was not specifically directed toward epidermal cell conjugates. Thulin and Zacharian (1972) extended others’ earlier work on MIF-induced migration of human peripheral blood lymphocytes to a test for delayed contact hypersensitivity. None of these approaches has yet been developed as an in vitro predictive test, but work is progressing. Milner (1983) has published a review of the history and state of this field. Any alternative (in vitro or in vivo) test for sensitization will need to be evaluated against a battery of “known” compounds. The Consumer Product Safety Commission in 1977 proposed such a battery, which is shown in Table 9.15. Gad (1987, 1988) have published comparative data on multiple animal and human test system data for some 72 materials. Such a database should be broadened and developed for other test systems.

APPROACHES Suggested Approaches to Testing As outlined above, there are numerous assays available to assess the various end points that are relevant to immunotoxicity. Early in the development process, a new compound should be evaluated with regard to various factors that may flag it as a potential immunotoxin, including chemical, structural, or physicochemical properties (e.g., photoallergin) and therapeutic class (i.e., immunomodulators, anti-inflammatories, and antimetabolites). Compounds from therapeutic or structural classes that are known to be potential immunotoxins or immunomodulators should be evaluated for the effects in question on a case-by-case basis. With the exception of immunomodulators, protein products, and products of biotechnology, the majority of pharmaceuticals can be assessed for most forms of immunotoxicity during routine preclinical toxicity tests. In general, a well-conducted preclinical toxicity study can detect most serious immunotoxins in the form of altered clinical, hematologic, or histological end points. For example, possible effects on humoral immunity may be indicated from clinical observation of gastrointestinal or respiratory pathology, changes in serum total protein and globulin, and by histological changes in lymphoid cellularity.

Likewise, effects on the cell-mediated response may be observed as increases in infections and tumor incidences, and by changes in the T-cell compartments of lymphoid tissues. In the case of immunosuppressive drugs such as cyclophosphamide and cyclosporin A, the immune effects seen in rodents are similar to those observed in the clinic (Dean et al., 1987). If perturbations are observed in any hematologic or histopathologic indicators of immunotoxicity, it is then prudent to follow up these findings with one or more of the following: l Use of special immunochemical and cytological assays that can be performed retrospectively on samples taken from the animals in question. ● Use of more specific in vitro assays to further assess effects on the pertinent target system and potential mechanism of activity. ● Use of more specific in vivo and ex vivo assays to determine toxicological significance. ● Inclusion of additional nonroutine parameters for immunotoxicity assessment in subsequent (longer-term) toxicity assays. Can also include additional satellite groups for functional tests that may require coadministration of adjuvants, pathogens, or tumor cells.

Suggested Approaches to Evaluation of Results Several rodent toxicity studies have shown impaired host resistance to infectious agents or tumor cells at exposure levels of drugs that did not cause overt signs of toxicity (Vos, 1977; Dean et al., 1982). One serious limitation to the incorporation of specific immunotoxicological evaluations in to general use in safety assessment for pharmaceuticals is a lack of clarity in how to evaluate and use such findings. This problem is true for all new diagnostic techniques in medicine and for all the new and more sensitive tools designed to evaluate specific target organ toxicities. Ultimately, as we have more experience and a reliable data base that allows us to correlate laboratory findings with clinical experience, the required course of action will become clearer. However, some general suggestions and guidance can be offered. ●

●

●

First, it is generally agreed that adverse effects observed above a certain level of severity should be given the same importance as any other life-threatening events when assessing biological significance. These are effects that are so severe that they are detected as part of the routine evaluations made in safety assessment studies. Such findings may include death, severe weight loss, early appearance of tumors, and the like. Findings such as significantly increased mortalities in a host-resistance assay would also fit into this category. Second, there are specific end point assays for which an adverse outcome clearly dictates the action to be taken. These end points include either immediate or delayed hypersensitivity reactions, because once the individual is sensitized, a dose–response relationship may not apply. Third, as with most toxicological effects, toxic effects to the immune system are dependent upon dose to the target site. The dose–response curve can be used to determine no-effect and low-effect levels for immunotoxicity. These levels can then be compared to the therapeutic levels to assess whether there is an adequate margin of safety for humans.

Chapter 9:

If we consider both the specific immunotoxicity assays surveyed earlier in the chapter and the arrays of end points evaluated in traditional toxicology studies, which may be indicative of an immune system effect, these guidelines leave many potential questions unanswered. As additional data on individual end points indicative of immune system responses are collected, the pharmaceutical toxicologist is challenged with various issues regarding assay interpretation and relevance to proposed (or future) clinical trials. For example, what do significant, but non-life-threatening, decreases in antibody response, lymphocyte numbers, macrophage functions, or host resistance in an animal mean about the clinic use of a drug in a patient? The intended patient population is clearly relevant here—if the disease is one in which the immune system is already challenged or incorrectly modulated, any immune system effect other than an intended one should be avoided. There are several additional considerations and questions that should be answered when evaluating the biological and clinical significance of a statistically significant immune response: 1. Is there a dose response? The dose response should be evaluated as a dose-related trend in both incidence and severity of the response. If there is a dose-related response, is the lowest dose (preferably plasma level) at which the effect is seen near or below the target clinical dose (plasma level), and is there an adequate therapeutic margin of safety? 2. Does the finding stand alone? Is a change observed in only one parameter, or are there correlated findings that suggest a generalized, biologically significant effect? For example, are there changes in lymph node and spleen weights and morphological changes in these tissues to accompany changes in lymphocyte numbers? 3. Is the effect a measure of function or a single end point measurement? Functional measures such as host resistance or phagocytosis involve multiple cells and immunocomponents and, therefore, are considered to be more biologically relevant than a significant change in a single end point measurement (e.g., T-cell number). 4. Is the effect reversible? Reversibility of a response is dependent on the drug itself, exposure levels/duration, and factors related to the test animal (metabolic capability, genetic susceptibility, etc.). Most effects produced by immunosuppressive drugs have been shown to be reversible after cessation of therapy, such as those produced during cancer chemotherapy. However, if a tumor develops before the immune system is restored, the effect is not reversible, as is the case of secondary tumors related to chemotherapy. 5. Is there sufficient systematic toxicity data available at levels that demonstrate adequate exposure? If a study was designed such that there was insufficient exposure or duration of exposure to potential lymphoid target tissues, the test protocol may not be adequate to demonstrate an adverse effect. In general, a well-conducted long-term study in two species, with no indication of immunotoxicity, based on the considerations outlined above, should be adequate to evaluate the potential for drug-induced immunotoxicity. If the results from these studies do not produce evidence of immune-specific toxicity after examination of standard and/or additional hematologic, serum chemical, and histopathologic parameters, then additional testing should not be indicated. However, if there are structure-activity

Immunotoxicology

127

considerations that may indicate a potential for concern, or if significant abnormalities are observed that cannot be clearly attributed to other toxicities, then it is important to perform additional tests to fully assess the biological significance of the findings.

PROBLEMS AND FUTURE DIRECTIONS There are some very pressing problems for immunotoxicology, particularly in the context of pharmaceuticals and biological therapeutics and the assessment of their safety. Unlike industrial chemicals, environmental agents, or agricultural chemicals, pharmaceutical products are intended for human exposure, are usually systemically absorbed, and have intentionally biological effects on man—some of which are intentionally immunomodulating (interleukins, growth factors) or immunotoxic (cyclosporin, cyclophosphamide). Data interpretation. The first major issue was presented and explored in the preceding section. This is how to evaluate and utilize the entire range of data that current immunotoxicological methodologies provide to determine the potential for immunotoxicity, and how to interpret the biological significance of minor findings. Appropriate animal models. As previously addressed, most routine preclinical toxicology tests are performed with rats and dogs; therefore, toxicity, pharmacokinetic, and pharmacology data are most abundant for these species. However, most immunological parameters are best characterized and validated with mice. In addition, the NTP test battery was developed for the mouse, and some of these assays cannot be readily transferred to the rat. Over the last few years, several laboratories have begun adapting tests to both the rat and dog (Bloom et al., 1985a, b; Thiem et al., 1988); however, efforts need to continue along these lines to further our understanding of the immune responses in these species and how they correlate with other animal models and man. Indirect immunotoxic effects. A problem related to data interpretation is how to distinguish secondary effects that may indirectly result in immunotoxicity from the primary effects of immunotoxicity in preclinic toxicity studies. Various factors may produce pathology similar to that of an immunotoxin, including: ●

●

●

Stress in a chronically ill animal as related to general toxicity, such as lung or liver damage, can result in immune suppression. Malnutrition in animals with drug-induced anorexia or malabsorption can trigger immune suppression. Infections and/or parasites may also modulate immune parameters.

These indirect factors must be systematically ruled out, and additional mechanistic studies may be necessary to address this problem. The potential for some indirect effects may be assessed through histopathologic evaluation of endocrine organs such as the adrenals and the pituitary. Hypersensitivity tests. Probably the largest immunotoxicity concern in clinical studies is unexpected hypersensitivity reactions. While the available guinea pig- and mouse-based tests for delayed contact hypersensitivity resulting from dermal exposure are generally good predictors, there are currently no well-validated models for either immediate or delayed hypersensitivity responses resulting from either oral ingestion or parenteral administration. Yet these two

128

Safety Evaluation of Medical Devices

situations are the largest single cause for discontinuing clinical trials. One assay that may hold some promise for delayed hypersensitivity is an adoptive transfer-popliteal lymph node assay (Klinkhammer et al., 1988). This assay, based on the techniques previously described for the popliteal lymph node assay, allows assessment of hypersensitivity following systemic exposure of the drug. Donor mice are first injected with drug for five consecutive days. After a four-week latency period, potentially sensitized T cells obtained from the spleen are injected into the footpad of a syngeneic mouse together with a subcutaneous challenge dose of the drug. Two to five days after the cell transfer, the popliteal lymph nodes are measured and observed for evidence of a response (enlargement). Once this assay is validated, it should allow for a more relevant assessment of hypersensitivity for drugs that are administered systemically (Gleichman et al., 1989). Autoimmunity. Traditional methods for assessing immunotoxicity as part of routine preclinical toxicity tests are primarily geared toward the detection of immunosuppressive effects. Although it is possible to incorporate clinical methods for detecting immune complexes and autoantibodies into the preclinical test protocols, the significance of adverse findings is ambiguous. Since these effects have a genetic component to their expression, the relevance of findings in animals is of questionable significance, particularly since these findings in the clinic do not always correlate with pathological effects. Functional reserve capacity. As previously discussed, the immune system has a tremendous reserve capacity that offers several levels of protection and back-ups to the primary response. As a consequence, this functional reserve can allow biologically significant, immunotoxic insults to occur without the appearance of morphological changes. Furthermore, adverse effects may remain subclinical until the organism is subjected to undue stress or subsequent challenge. Thus, there is some concern that routine immunopathological assessments by themselves may not be sufficiently sensitive to detect all immunotoxins, particularly when testing is conducted in a relatively pathogen-free, stress-free laboratory environment. Significance of minor perturbations. Although the immune system has a well-developed reserve capacity, some of these systems may act synergistically rather than independently. For instance, a macrophage can recognize and kill bacteria coated with antibodies more effectively than can either the macrophage or antibodies alone. Thus, even minor deficiencies and impairments may have some impact on the organism’s ability to fend off infection or tumors, particularly if the organism is very young, old, ill, stressed, genetically predisposed to certain cancers, or otherwise immunocompromised. These considerations lead to some additional questions that must be addressed: ●

●

●

●

What level of immunosuppression will predispose healthy or immunocompromised individuals to increased risk of infection or tumors? What slight disturbances or immunosuppression lead to a prolonged recovery from viral or bacterial infections? Will slight up-modulation for extended periods result in autoimmune diseases or increased susceptibility to allergy? Are individuals that are slightly immunosuppressed at higher risk of developing AIDS after exposure to HIV?

The blind men and the elephant—this parable applies all too well to evaluating the relevance of isolated measures

of effect on the immune system to truly adverse changes in patient health. We can measure with great sensitivity and precision changes in many components of the system. But we do not yet have sufficient understanding to know where the right line is that says it is an indication of an adverse effect on health.

REFERENCES Adams, D.O. (1983). The biology of the granuloma. In: H.L. Ioachim, (Ed.), Pathology of Granulomas. Raven Press, New York, NY, pp. 1–20. Anderson, J.M., Inflammatory response to implants (1988). Trans. Am. Soc. Artif. Intern. Organs, 34: 101–107. Arts, J.H.E., Droge, S.C.M., Bloksma, N. and Kuper, C.F. (1996). Local lymph node activation in rats after application of the sensitizers 2,4-dinitrichlorobenzene and trimellic anhydride. Fd. Chem. Toxicol., 34: 55–62. Asherson, G.L. and Ptak, W. (1968). Contact and delayed hypersensitivity in the mouse. I. Active sensitization and passive transfer. Immunology, 15: 405–416. Bach, F.H. and Voynow, N.K. (1966). One-way stimulation in mixed leukocyte cultures. Science, 153: 545–547. Bakke, O.M., Wardell, W.M., and Lasagna, L. (1984). Drug discontinuations in the United Kingdom and United States, 1964–1983: Issues of safety. Clin. Pharmacol. Therapy, 35: 559–567. Basketter, D.A., Scholes, E.W., Kimber, I., Botham, P.A., Hilton, J., Miller, K., Robbins, M.C., Harrison, P.T.C. and Waite, S.J. (1991). Interlaboratory evaluation of the local lymph node assay with 25 chemicals and comparison with guinea pig test data. Toxicol. Methods, 1: 30–43. Basketter, D.A. and Scholes, E.W. (1991). Comparison of the local lymph node assay with the guinea pig maximization test for the detection of a range of contact allergens. Fd. Chem. Toxic., 30: 65–69. Battisto, J.R., Claman H.N. and Scott, D.W. (1983). Immunological tolerance to self and nonself. Ann. N.Y. Acad. Sci., 392: 1–433. Bellon, B., Capron, M., Druet, E., Verroust, P., Vial, M.-C., Girard, J.F., Foidart, J.M., Mahieu, P. and Druet, P. (1982). Mercuric chloride induced autoimmune disease in Brown-Norway rats: Sequential search for anti-basement membrane antibodies and circulating immune complexes. Eur. J. Clin. Invest., 12: 127–133. Bick, P.H. (1985). The immune system: organization and function. In: J.H. Dean, A.E. Munson, M.I. Luster, and H. Amos (Eds.), Immunotoxicology and Immunopharmacology. Raven Press, New York, pp. 1–10. Bigazzi, P.E. (1985). Mechanisms of chemical-induced autoimmunity. In: J.H. Dean, M.I. Luster, A.E. Munson, and H. Amos (Eds.), Immunotoxicology and Immunopharmacology. Raven Press, New York, p. 277. Bigazzi, P.E. (1988). Autoimmunity induced by chemicals. Clin. Toxicol., 26: 125–156. Black, J. (1981). Rating system for tissues at animal implant sites. Biological Performance of Materials, Fundamentals of Biocompatibility. Marcel Dekker, New York, NY, pp. 220, Appendix 2. Bloom, J.C., Thiem, P.A., Sellers, T.S., Deldar, A. and Lewis, H.B. (1985a). Cephalosporin-induced immune cytopenia in the dog— Demonstration of cell associated antibodies. Blood, 66: 1232. Bloom, J.C., Blackmer, S.A., Bugelski, P.J., Sowinski, J.M. and Saunders, L.Z. (1985b). Gold-included immune thrombocytopenia in the dog. Vet. Pathol., 22: 492–499. Bloom, J.C., Thiem, P.A. and Morgan D.G. (1987). The role of conventional pathology and toxicology in evaluating the immunotoxic potential of xenobiotics. Toxicol. Path., 15: 283–293. Buehler, E.V. (1964). A new method for detecting potential sensitizers using the guinea pig. Toxicol. Appl. Pharmacol., 6: 341. Buehler, E.V. (1965). Delayed contact hypersensitivity in the guinea pig. Arch. Dermatol., 91: 171–177. Burkitt, D.S., Barwell, N.J. and Wilson, A.G.M. (1986). Urethral catheter structures. Lancet, 1: 688. Burnet, F.M. (1970). The concept of immunological surveillance. Progr. Exper. Tumor Res., 13: 1–27.

Chapter 9: CDRH (1997). Immunotoxicology Testing Framework, Draft Document. Christensen, O.B., Christensen, M.B. and Maibach, H.I. (1984). Effect of vehicle on elicitation of DNCB contact allergy in the guinea pig, Contact Dermatitis, 10: 166–169. Clark, W.R. (1983). The Experimental Foundations of Modern Immunology, Second edn. John Wiley & Sons, New York, pp. 1–453. Coombs, R.R.A. and Gell, P.G.H. (1975). Classification of allergic reactions responsible for clinical hypersensitivity and disease. In: P.G.H. Gell, R.A.A., Coombs, and D.J. Lachman (Eds.), Clinical Aspects of Immunology. Blackwell Scientific Publications, Oxford, pp. 761. Cromwell, O., Durham, S.R., Shaw, R.J., Mackay, J.A. and Kay, A.B. (1986). Provocation tests and measurements of mediators from mast cells and basophils in asthma and allergic rhinitis. In: D.M. Weir, L.A. Herzenberg, C. Blackwell (Eds.), Handbook of Experimental Immunology, Fourth edn. Blackwell, Oxford, pp. 127.1–127-51. Cronin, E. and Agrup, G. (1970). Contact Dermatitis X. Brit. J. Dermatol., 82: 428–433. Crowle, A.J. (1959a). Delayed hypersensitivity in mice: its detection by skin tests and its passive transfer. Science, 130: 159. Crowle, A.J. (1959b). Delayed hypersensitivity in several strains of mice studied with six different tests. J. Allergy, 30: 442–459. Dean, J.H., Luster, M.I. and Boorman, G.A. (1982). Immunotoxicology. In: P. Sirois and M. Rola-Pleszczynski (Eds.), Immunopharmacology. Elsevier Biomedical Press, Amsterdam, pp. 349–397. Dean, J.H., Boorman, G.A., Luster, M.I., Adkins, B.J., Lauer, L.D. and Adams, D.O. (1984). Effect of agents of environmental concern on macrophage functions. In: A. Volkman (Ed.), Mononuclear Phagocyte Biology. Dekker, New York, pp. 473–485. Dean, J.H., Thurmond, L.M., Lauer, L.D. and House, R.V. (1987). Comparative toxicology and correlative immunotoxicology in rodents. In: E.J., Burger, R.G. Tardiff and J.A. Bellanti (Eds.), Environmental Chemical Exposure and Immune System Integrity, Princeton Scientific Publishing Company, Princeton, pp. 265–271. Descotes, G. and Mazue, G. (1987). Immunotoxicology. Advances in Veterinary Science and Comparative Medicine, 31: 95–119. DeSousa, M.A.B. and Parrott, D.M.V. (1969). Induction and recall in contact sensitivity, changes in skin and draining lymph nodes of intact and thymectomized mice. J. Exp. Med., 130: 671–686. DeSwarte, R.D. (1986). Drug allergy. An overview. Clin. Rev. Allergy, 4: 143–169. DeWeck, A.L. (1983). Immunopathological mechanisms and clinical aspects of allergic reactions to drugs. In: A.L. deWeck and H. Burdgaard (Eds.), Handbook of Experimental Pharmacology: Allergic Reactions to Drugs. Springer-Verlag, New York, pp. 75–133. Dighiero, G., Lymberi, J., Marie, J.C., Rouyse, S., Butler-Browne, G.S., Whalen, R.G. and Avrameas, S. (1983). Murine hybridomas secreting natural monoclonal antibodies reacting with self antigens. J. Immunol., 135: 2267–2271. Draize, J.H. (1959). The Appraisal of Chemicals in Foods, Drugs and Cosmetics, Association of Food and Drug Officials of the U.S., Austin, TX, pp. 36–45. Draize, J.H., Woodard, G. and Calvery, H.O. (1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucus membranes. J. Pharmacol. Exp. Ther., 82: 377–390. Druet, P., Bernard, A., Hirsch, F., Weening, J.J., Gengoux, P. Mahieu, P. and Brikenland, S. (1982). Immunologically mediated glomerulonephritis induced by heavy metals. Arch. Toxicol., 50: 187–194. Duke, S.S., Schook, L.B. and Holsapple, M.P. (1985). Effects of N-nitrosodimethylamine on tumor susceptibility. J. Leukocyte Biol., 37: 383–394. Exon, J.H., Koller, L.D., Talcott, P.A., O’Reilly, C.A. and Henningsen, G.M. (1986). Immunotoxicity testing: An economical multiple assay approach. Fund. Appl. Toxicol., 7: 387–397. FDA (Food and Drug Administration), Center for Safety and Applied Nutrition. (1993). Draft: Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food. Federal Register, 58: 10536. Gad, S.C. (1987). Scheme for the ranking and prediction of relative potencies of dermal sensitizers, based on data from several test systems. Toxicologist, A343.

Immunotoxicology

129

Gad, S.C. (1988). A Scheme for the Ranking and Prediction of Relative Potencies of Dermal Sensitizers Based on Data from Several Test Systems. J. Appl. Toxicol., 8: 301–312. Gad, S.C., Darr, R.W., Dobbs, D.W., Dunn, B.J., Reilly, C. and Walsh, R.D. (1986). Comparison of the potency of 52 dermal sensitizers in the Mouse Ear Swelling Test (MEST), presented at SOT Meetings, March 1986. Gad, S.C., Dobbs, D.W., Dunn, B.J., Reilly, C. and Walsh, R.D. (1985a). Elucidation of the delayed contact sensitization response to Croton Oil, presented at the American College of Toxicology in Washington, DC, November 1985. Gad, S.C., Dobbs, D.W., Dunn, B.J., Reilly, C. and Walsh, R.D. (1987). Development, validation and transfer of a new test system technology in toxicology. In: A.M. Goldberg (Ed.), New Test Systems in Toxicology, Vol. 5. Mary Ann Liebert, New York, pp. 275–292. Gad, S.C., Dunn, B.J. and Dobb, D.W. (1985b). Development of alternative dermal sensitization test: Mouse Ear Swelling Test (MEST). In: A.M. Golberg (Ed.), In Vitro Toxicology. Proceedings of 1984 Johns Hopkins Symposium, pp. 539–551. Gad S.C., Dunn, B.J., Dobbs, D.W. and Walsh, R.D. (1986). Development and validation of an alternative dermal sensitization test: the Mouse Ear Swelling Test (MEST). Tox. and Appl. Pharmacol., 84: 93–114. Garratlz, G. and Petz, L.D. (1975). Drug-induced immune hemolytic anemia. Am. I. Med., 58: 398–407. Gibson, G.G., Hubbard, R., and Parke, D.U. (1983) Immunotoxicology, Academic Press, San Diego, CA. Gleichmann, H. (1989). Testing the sensitization of T cells to chemicals. From murine graft-versus host (GVH) reactions to chemical-induced GVH-like immunological diseases. In: M.E. Kammueller, N. Bloksma, and W. Seinen (Eds.), Autoimmunity and Toxicology. Elsevier, Amsterdam, pp. 263–385. Gleichmann, E., Kimber, I. and Purchase, I.F.H. (1989). Immunotoxicology: suppressive and stimulatory effects of drugs and environmental chemicals on the immune system. Arch. Toxicol., 63: 257–273. Godfrey, H.P. and Gell, P.G.H. (1978). Cellular and molecular events in the delayed-onset hypersensitivities. Reviews in Physiological and Biochemical Pharmacology, 84: 2. Golub, E.S., and Green, D.R. (1991). Immunology: a Synthesis. Sinauer, Sunderland, Massachusetts, pp. 1–744. Guillaume, J.C., Roujeau, J.C. and Touraine, R. (1984). Purine analogs as immunomodulators. In: Y. Yamamura and T. Tada (Ed.), Progress in Immunology IV. London: Academic Press, pp. 1393–1407. Guillot, J.P. and Gonnet, J.F. (1985). The Epicutaneous Maximization Test. Curr. Probl. Derm. 14: 220–247. Harmon, W.E., Parkman, R., Gavin, P.T., Grupe, W.E., Ingelfunger, J.R., Yunis, E.J. and Levey, R.H. (1982). Comparison of cell-mediated lympholysis and mixed lymphocyte culture in the immunologic evaluation for renal transplantation. J. Immunol., 129: 1573–1577. Hassman, R.A., Lazarus, J.H., Dieguez, C., Weetman, A.P., Hall, R. and McGregor, A.M. (1985). The influence of lithium chloride on experimental autoimmune thyroid disease. Clin. Exp. Immunol., 61: 49–57. Hinton, D.M. (1992). Testing guidelines for evaluation of the immunotoxic potential of direct food additives. Critical Reviews in Food Science and Nutrition, 32: 173–190. Hutchings, P., Nador, D. and Cooke, A. (1985). Effects of low doses of cyclophosphamide and low doses of irradiation on the regulation of induced erythrocyte autoantibodies in mice. Immunol., 54: 97–104. Idsøe, O., Guthe, T., Willcox, R.R. and DeWeck, A.L. (1968). Nature and extent of penicillin side-reactions, with particular reference to fatalities from anaphylactic shock. Bull. WHO, 38: 159–188. Jerne, N.K. and Nordin, A.A. (1963). Plaque formation in agar by single antibody-producing cells. Science, 140: 405. Kimber, I. and Dearman, R.J. (1994). Immune responses to contact and respiratory allergens. In: J.H. Dean, M.I. Luster, E.A. Munson and I. Kimber (Eds.), Immunotoxicology and Immunopharmacology. Raven Press, New York, pp. 663–679.

130

Safety Evaluation of Medical Devices

Kimber, I., Mitchell, J.A. and Griffin, A.C. (1986). Development of an immurine local lymph node assay for the determination of sensitizing potential. Food and Chemical Toxicology, 24: 585–586. Kimber, I. and Weisenberger, C. (1989). A murine local lymph node assay for the identification of contact allergens: Assay development and results of an initial validation study. Arch. Toxicol., 63: 274–282. Klaus, G.G.B. and Hawrylowicz, C.N.M. (1984). Cell-cycle control in lymphocyte stimulation. Immunol. Today, 5: 15–19. Kleinman, S., Nelson, R., Smith, L. and Goldfinger, D. (1984). Positive direct antiglobulin tests and immune hemolytic anemia in patients receiving procainamide. New Engl. J. Med. 311: 809–812. Kligman, A.M. (1966). The identification of contact allergens by human assay. III. The maximization test. A procedure for screening and rating contact sensitizers. J. Invest. Dermatol., 47: 393–409. Klinkhammer, C., Popowa, P. and Gleichmann, H. (1988). Specific immunity to the diabetogen streptozotocin: cellular requirements for induction of lymphoproliferation. Diabetes, 37: 74–80. Koller, L.D. (1987). Immunotoxicology today. Toxicol. Path., 15: 346–351. Kumagai, Y., Shiokawa, Y., Medsger, T.A.J. and Rodnan, G.P. (1984). Clinical spectrum of connective tissue disease after cosmetic surgery. Observations of eighteen patients and a review of the Japanese literature. Arthr. Rheum., 27: 1–12. Landsteiner, K. and Chase, M.W. (1942). Experiments on transfer of cutaneous sensitivity to simple chemical compounds. Proc. Soc. Exp. Biol. Med., 49: 288–390. Landsteiner, K. and Jacobs, J. (1935). Studies on sensitization of animals with simple chemical compounds. J. Exp. Med., 61: 643–656. Laskin, D.L., Laskin, J.D., Weinstein, I.B., and Carchman, R.A. (1981). Induction of chemotaxis in mouse peritoneal macrophages by phorbolester tumor promoters. Cancer Res., 41: 1923. Logue, G.L., Boyd, A.E. and Rosse, W.F. (1970). Chlorpropamideinduced immune hemolytic anemia. New Engl. J. Med., 283: 900–904. Loveless, S.E., Ladics, G.S., Greberick, G.F., Ryan, C.A., Basketter, D.A., Scholes, E.W., House, R.V., Hilton, J., Dearman, R.J. and Kimber, I. (1996). Further evaluations of the local lymph node assay in the final phase of the international collaborative trial. Toxicology, 108: 141–152. Luster, M.I., Munson, A.E., Thomas, P.T., Holsapple, M.P., Fenters, J.D., White, Jr., K.L., Lauer, L.D., Germolec, D.R., Rosenthal, G.L. and Dean, J.H. (1988). Development of a testing battery to assess chemical-induced immunotoxicity: National Toxicology Program’s guidelines for immunotoxicity evaluation in mice. Fund. Appl. Toxicol., 10: 2–19. Luster, M.I., Pait, D.G., Portier, C., Rosenthal, G.J., Dermolec, D.R., Comment, C.E., Munson, A.E., White, K. and Pollock, P. (1992a). Qualitative and quantitative experimental models to aid in risk assessment for immunotoxicology. Toxicol. Lett., 64/65: 71–78. Luster, M.I., Portier, C., Pait, D.G., White, K.L., Jr., Gennings, C., Munson, A.E. and Rosenthal, G.J. (1992b). Risk assessment in Immunotoxicology. I. Sensitivity and predictability of immune tests. Fund. Appl. Toxicol., 18: 200–210. Magnusson, B. (1975). The relevance of results obtained with the guinea pig maximization test. In: H. Maibach (Ed.), Animal Models in Dermatology. Churchill Livingstone, Edinburgh, pp. 76–83. Magnusson, B. and Kligman, A.M. (1969). The identification of contact allergens by animal assay. The guinea pig maximization test. J. Invest Dermatol., 52: 268–276. Magnusson, B. and Kligman, A.M. (1970). Allergic Contact Dermatitis in the Guinea Pig: Identification of Contact Allergens, Chapter 8. Thomas, Springfield, Illinois. Maguire, H.C. and Chase, M.W. (1967). Exaggerated delayed-type hypersensitivity to simple chemical allergens in the guinea pig. J. Invest. Dermatol., 49: 460–468. Male. D., Champion, B. and Cooke, A. (1982). Advanced Immunology. J.B. Lippincott, Philadelphia.

Marchant, R.E., Miller, K.M., Hiltner, A. and Anderson, J.M. (1985). Selected aspects of cell and molecular biology of in vivo biocompatibility. In: S.W. Shalaby, A.S. Hoffman, B.D. Ratner and T.A. Horbett (Eds.), Polymers as Biomaterials. Plenum Press, New York, pp. 209–223. Marzulli, F.N. and Maibach, H.L. (1996). Dermatotoxicity, 5th Edn. Taylor and Francis, Philadelphia, PA. Matory, Y.L., Chang, A.E., Lipford, E.H., Braziel, R., Hyatt, C.L., McDonald, H.D. and Rosenber, S.A. (1985). Toxicity of recombinant human interleukin-2 in rats following intravenous infusion. J. Biol. Response Mod., 4: 377–390. Maurer, T., Thomann, P., Weirich, E.G. and Hess, R. (1975). The optimization test in the guinea pig. Agents and Actions, 5: 174–179. Maurer, T., Weirich, E.G. and Hess, R. (1980). The optimization test in the guinea pig in relation to other predictive sensitization methods. Toxicology, 15: 163–171. Merluzzi, V.J. (1985). Comparison of murine lymphokine, activated killer cells, natural killer cells, and cytotoxic T lymphocytes. Cell Immunol., 95: 95–104. Miller, A.E., Jr. and Levis, W.R. (1973). Studies on the contact sensitization of man with simple chemicals, I. Specific lymphocyte transformation in response to dinitrochlorobenzene sensitization. J. Invest. Dermatol., 61: 261–269. Milner, J.E. (1970). In vitro lymphocyte responses in contact hypersensitivity. J. Invest. Dermatol., 55: 34–38. Milner, J.E. (1971). In vitro lymphocyte responses in contact hypersensitivity II. J. Invest. Dermatol., 56: 349–352. Milner, J.E. (1983). In vitro tests for delayed skin hypersensitivity: lymphokine production in allergic contact dermatisis. In: F.N. Marzulli and H.D. Maibach (Eds.), Dermatotoxicology. Hemisphere Publishing, New York, pp. 185–192. NAS Publication 1138 (1977). Principles and Procedures for Evaluating the Toxicity of Household Substances. Prepared for the Consumer Product Safety Commission, National Academy of Sciences, Washington, D.C., pp. 36–39. Paranjpe, M.S. and Boone, C.W. (1972). Delayed hypersensitivity to simianvirus 40 tumor cells in BALB/c mice demonstrated by a radioisotopic footpad assay. J. Natl. Cancer Inst., 48: 563. Pastan, I., Willingham, M.C. and FitzGerald, D.J. (1986). Immunotoxins. Cell, 47: 641–648. Penn, I. (1977). Development of cancer as a complication of clinical transplantation. Transplant. Proc., 9: 1121–1127. Pennisi, I. (1996). Teetering on the brink of danger. Science, 271: 1665–1667. Riccardi, C., Puccetti, P., Santoni, A. and Herberman, R.B. (1979). Rapid in vivo assay of mouse natural killer cell activity. J. Natl. Cancer Inst., 63: 1041–1045. Roitt, I., Brostoff, J. and Male, D. (1985). Adaptive and innate immunity. In: I.J. Roitt, Brostoff and D. Mabe (Eds.), Immunology. C. V. Mosby, St. Louis, MO, pp. 1.1–1.10. Russel, A.S. (1981). Drug-induced autoimmune disease. Clin. Immun. Allergy, 1: 57. Salthouse, T.N. (1982). In: D.F. Williams (Ed.), Biocompatibility in Clinical Practice. CRC Press, Boca Raton, FL, pp. 18–24. Spreafico, F. (1988). Immunotoxicology in 1987: Problems and challenges. Fund. Clin. Pharmacol., 2: 353–367. Sullivan, J. (1989). Immunological alterations and chemical exposure. Clin. Toxicol., 27: 311–343. Tabo, J.D. and Paul, W.E. (1973). Functional heterogeneity of murine lymphoid cells. III. Differential responsiveness of T cells to phytohemaglutinin and concanavalin A for T cell subsets. J. Immunol., 110: 362–369. Talal, N. (1987). Autoimmune mechanisms in patients and animal models. Toxicol. Path., 15: 272–275. TCSA, Federal Register, Vol. 44, No. 145, July 26 (1979). Part IV, proposed EPA Test Standards for Toxic Substances Control Act Test Rules, Part 772.112–26. Thiem, P.A., Halper, L.K. and Bloom, J.C. (1988). Techniques for assessing canine mononuclear phagocyte function as part of an immunotoxicologic evaluation. Int. J. Immunopharm., 10: 765–771.

Chapter 9: Thorne, P.S., Hillebrand, J.A., Lewis, G.R. and Karol, M.H. (1987). Contact sensitivity by diisocyanates: potencies and cross-reactivities. Tox. Appl. Pharmacol., 87: 155–165. Thulin, H. and Zacharian, H. (1972). The leukocyte migration test in chromium hypersensitivity. J. Invest. Dermatol., 58: 55–58. Ueda, S., Wakahim, Y., Takei, I., Mori, T., and Lesato, K. (1980). Autologous immune complex nephritis in gold injected guinea pigs. Nippon Jinzo Gakkai Shi, 22: 1221–1230. Unanue, E.R. (1984). The immune granulomas. In: E.R. Unanue, and B. Benaceraf, (Eds.), Textbook of Immunology. Williams and Wilkins, Baltimore, MD, pp. 297–306. Volkman, A. (1984). Mononuclear Phagocyte Function. Marcel Dekker, New York. Vos, J.G. (1977). Immune suppression as related to toxicology. CRC Critical Reviews in Toxicology, 5: 67. Vos, J., Van Loveren, H., Weester, P. and Vathaak, D. (1989). Toxic effects of environmental chemicals on the immune system. TIPS, 10: 289–292.

Immunotoxicology

131

Weening, J.J., Fleuren, G.J., and Hoedemaeker, J. (1978). Demonstration of antinuclear antibodies in mercuric chloride-induced glomerulopathy in the rat. Lab. Invest., 39: 405–411. Weigle, W.O. (1980). Analysis of autoimmunity through experimental models of thyroditis and allergic encephalomyelitis. Adv. Immunol., 30: 159–275. White, K.L., Jr., Sanders, V.M., Barnes, D.W., Shopp, J.G.M. and Munson, A.E. (1985). Immunotoxicological investigations in the mouse: General approach and methods. Drug. Chem. Toxicol., 8: 299–331. Woodward, S.C. and Salthouse, T.N. (1986). The tissue response to implants and its evaluation by light microscopy. In: A.F. von Recum (Ed.), Handbook of Biomaterials. Macmillan Publishing, New York, pp. 364–379. Yoshida, S., Golub, M.S. and Gershwin, M.E. (1989). Immunological aspects of toxicology: Premises not promises. Reg. Toxicol. Pharm., 9: 56–80.

10 Implantation Biology and Studies

Implantation studies are a type of assay unique to medical device and biomaterials, having been specifically devised for those situations where an exogenous (and usually manmade) construct or material is enclosed in the body or partially entered into it by a breached surface. It is intended to assess the effects of devices (usually polymers or elastomers) which are in direct contact with living tissue (not including the skin). The effects of concern may be either short or long term, with a range of responses over the course of the host body and the device interacting with each other for various indicators of local tissue tolerance. Longer term studies are conducted for long term implants, and focus more on broader systemic effects and potential carcinogenicity. The whole field of implantation biology (which studies these interactions) is an extremely complete and active one (Greco, 1994; Black, 2000; Guelcher and Hollinger, 2006).

the animals on both sides of the spinal column. Remove loose hair by means of vacuum. Swab the skin lightly with diluted alcohol and dry the skin prior to injection. Implant four strips of the Sample into the paravertabral muscle on one side of the spine of each of two rabbits, 2.5–5 cm from the midline and parallel to the spinal column, and about 2.5 cm apart from each other. In a similar fashion implant two strips of USP Negative Control Plastic RS in the opposite muscle of each animal. Insert a sterile stylet into the needle to hold the implant strip in the tissue while withdrawing the needle. If excessive bleeding is observed after implantation of a strip, place a duplicate strip at another site. Keep the animals for a period of not less than 120 hours, and sacrifice them at the end of the observation period by administering an overdose of an anesthetic agent or other suitable agents. Allow sufficient time to elapse for the tissue to be cut without bleeding. Examine macroscopically the area of the tissue surrounding the center portion of each implant strip. Use a magnifying lens if necessary. The tissue immediately surrounding the USP Negative Control Plastic RS strips appears normal and entirely free from hemorrhage, film, or encapsulation. The requirements of the test are met if, in each rabbit, the reaction to not more than 1 of the 4 Sample strips is significantly greater than that to the strips of USP Negative Control Plastic RS.

USP IMPLANTATION TEST The USP (USP, 2007) test is designed to evaluate relatively short term exposures, and it is the normative reference method for both the FDA and ISO (10993-6). The implantation test is designed for the evaluation of plastic materials and other polymeric materials in direct contact with living tissue. Of importance are the proper preparation of the implant strips and their proper implantation under aseptic conditions. Prepare for implantation eight strips of the Sample and four strips of USP Negative Control Plastic RS. Each strip should measure not less than 10 × 1 mm. The edges of the strips should be as smooth as possible to avoid additional mechanical trauma upon implantation. Strips of the specified minimum size are implanted by means of a hypodermic needle (15- to 19-gauge) with intravenous point and a sterile trocar. Use either pre-sterilized needles into which the sterile plastic strips are aseptically inserted, or insert each clean strip into a needle, the cannula and hub of which are protected with an appropriate cover, and then subjected to the appropriate sterilization procedure. [Note: Allow for proper degassing if agents such as ethylene oxide are used.] Test Animal—Select healthy, adult rabbits weighing not less than 2.5 kg, and whose paravertebral muscles are sufficiently large in size to allow for implantation of the test strips. Do not use any muscular tissue other than the paravertebral site. The animals may be anesthetized with a commonly used anesthetic agent to a degree deep enough to prevent muscular movements, such as twitching. Procedure—Perform the test in a clean area. On the day of the test or up to 20 hours before testing, clip the fur of

Observation macroscopically hemorrhage necrosis discolorations infections encapsulation. Histopathology (not required, but generally conducted) connective tissue proliferation mononuclear cell infiltration polymorphonuclear cell infiltration muscle degeneration multinucleated giant cell infiltration. Table 10.1 summarizes the most common microscopic findings in an implantation study (Greco, 1994; von Recum, 1998).

BRITISH PHARMACOPOEIA Implantation tests in rabbits are also required for medical devices and plastics, but in this case to evaluate the local effects of direct contact between solid samples and muscle tissue. The British specification (British Standard 5736/2, 1981) is for

133

134

Safety Evaluation of Medical Devices

Table 10.1

Microscopic Examination Findings in Implant Studies

Cell Type

Implication

Granulocytes and neutrophils

Cells of acute inflammation: first line of defense against bacteria and absorbable materials. These cells attempt to engulf and digest materials. Their presence implies a bioreactive material. Cells of subacute inflammation: second line of defense and more resistant to damage by breakdown products. Their presence implies a prolongation of the effort to degrade the bioreactive material. Cells of chronic inflammation: giant cells result form the fusion of monocytes and macrophages. Their presence is indicative of bioreactive materials that are resistant to degradation. These are the cells that encapsulate and isolate foreign particles from surrounding tissue. Their presence implies that the body has surrendered in its effort to rid itself of the foreign material and has chosen to isolate instead. Large densities of lymphocytes are characteristic of immunologic injury. Their presence in large numbers implies an immunologic response to the material.

Monocytes and macrophages

Giant cells

Fibroblasts

Lymphocytes

ISO 10993 IMPLANTATION TEST The ISO 10993 mandated test is covered by Part 6 of the test guidelines, and is specifically intended to test for local effects after implantation. It can be performed for either short-term (from 1 up to 12 weeks) or long-term (from 12 to 104 weeks). The test specimen is implanted into a site and tissue appropriate for evaluation of the biological safety of the material. The implant is not intended to be subjected to mechanical or functional loading. The local effects are evaluated by a comparison of the tissue response caused by a test specimen to that caused by materials used in medical devices whose clinical acceptability has been established.

Preparation of Specimens for Implantation Solid Specimens (Excluding Powders) Physical characteristics (that is form, density, hardness, surface finish) can influence the character of the tissue response to the test material. Each implant shall be manufactured, processed, cleaned of contaminants, and sterilized by the method intended for the final product. After final preparation and sterilization, the implant specimens shall be handled in such a way as to ensure that they are not scratched, damaged, or contaminated in any way prior to or during insertion.

Non-Solid Specimens (Including Powders) medical devices intended for long-term implantation, such as hip prostheses, and for short-term use within the body or in contact with mucosal surfaces, such as urinary catheters. The USP specification (United States Pharmacopoeia, 2007) is for plastics and other polymers intended for fabricating containers or their accessories, and for use in medical devices, implants, and other materials. The British method is similar but specifies not fewer than three rabbits, additional implantation or positive control strips (such as tin-stabilized polyvinylchloride), a duration of 7 days, and histology of the sites if any macroscopic reaction to the test material is negative. It also advocates a less objective assessment of the results. The anesthetic suggested is pentobarbitone (pentobarbital) but the neuroleptic–analgesic combination of fluanisone and fentanyl citrate is preferable because if is safer in rabbits and reversible. One problem of the implantation test is a tendency for the strips to migrate from their implantation sites, even to subcutaneous positions, and this often prolongs the search for them. Nevertheless it is an effective detection system for toxic ingredients of solid materials which leach in contact with tissue fluid. It is important to recognize the microscopic effects of the standard negative control strips (additive-free polyethylene). These are typical of skeletal muscle in contact for a week with a foreign body and comprise mild mononuclear cell infiltration, multinucleated giant cell formation, fibroplasia, slight dystrophic calcification, muscle fiber atrophy, and centripetal migration of sarcolemmal nuclei. Also, traumatic hemorrhage is common. Positive reactions are similar but more pronounced and additionally include focal necrosis and education, particularly of heterophils. It is a useful test, not only for finished products, but also to identify unacceptable changes in formulation or manufacturing process such as the introduction of chlorinating cycles to remove bloom on latex catheters.

Non-solid specimens may be liquids, pastes, and particulates, as distinct from the materials covered otherwise. The components may be mixed before use (e.g., bone cements, dental materials), and set after varying time periods. The materials may be contained in tubes for the purpose of testing for local effects after implantation. Polyethylene (PE), polypropylene (PP), or polytetrafluoroethylene (PTEE) tubes are commonly used for this purpose. Prior to test the tubes shall be rinsed with 70% (v/v) ethanol and distilled water and sterilized by autoclaving or other appropriate methods relevant for clinical applications. Materials tested in their freshly mixed state shall be tested for microbiological contamination. Prepare the test material according to the manufacturer’s instructions and insert the material into the tube until level with the top. Exercise the utmost care to prevent contamination of the outer surface of the tube by the test material. Avoid entrapment of air in the tube and ensure that the end surfaces of the inserted material in the tube and the tube ends are smooth. Note—PE tubes may be deformed by autoclaving. It is difficult to section PTFE tubes in the microtome, and substitution by PE or PP tubes of the same dimensions may be preferable when the tubes are to remain in the tissue blocks during sectioning.

Control Specimens The size, shape, and especially the surface condition of the control(s) shall be as similar to that of the implant test specimens as is practically possible. When the test material is contained in a tube, the control shall be a rod of the same material as the tube and with the same diameter as the outer diameter of the tube. The control specimens shall be handled, cleaned, and sterilized in such a manner as to maintain them as acceptable and well characterized controls. Selection of control material(s) should be based on their established use in clinical applications similar to those

Chapter 10:

proposed for the candidate test material and is not otherwise restricted.

Animals and Tissues Select an animal species with due consideration of the size of the implant test specimens, the intended duration of the test in relation to the expected life-span of the animals, as well as the recognized species differences in biological response in both hard and soft tissues. For short-term testing in subcutaneous tissue and muscle, animals such as mice, rats, guineapigs, and rabbits are commonly used (Gad, 2006). Select one species among these. For long-term testing in subcutaneous tissue, muscle, and bone, animals such as rats, guinea-pigs, rabbits, dogs, sheep, goats, pigs, and other animals with a relatively long life expectancy are suitable. Select one species among these. The specimens of test and control materials shall be implanted under the same conditions in the same species of the same age, sex, and strain in corresponding anatomical sites. The number and size of implants inserted in an animal depends on the size of the species and the anatomical location of the implantation.

Test Periods The local tissue response to implanted materials is assessed in short-term tests up to 12 weeks and in long-term tests exceeding 12 weeks. Test periods are chosen to ascertain that a steady state has been reached with respect to biological response. The local biological response to implanted materials depends both on the properties of the materials and on the trauma of surgery. The tissue configuration found in the vicinity of an implant changes with the time elapsed after surgery. Usually, at one week observation periods, a high cell activity is found, followed by a transitional stage. In muscle and connective tissue, depending on the species, a steady state is seen in the cell population after 9–12 weeks. Implantation in bone tissues may need longer observation periods. Test periods shall be selected from those specified in Table 10.2 for short-term implantation, or from Table 10.3 for long-term implantation. Table 10.2 Selection of Test Periods for Short-Term Implantation in Subcutaneous Tissue and Muscle Implantation period (weeks) Species

1

3

Mice Rats Guinea pigs Rabbits

× × × ×

×

4 × × ×

9

Implantation Biology and Studies

135

Depending on the intended use of the test material, not all implantation periods may be necessary (see ISO 10993-1). An observation period of 104 weeks may be of interest in selected instances. The number of implants per animal and the number of animals per observation period are described in the appropriate sections below. A sufficient number of implants shall be inserted to ensure that the final number of specimens to be evaluated will give valid results.

Surgery Anesthetize the animals. Remove hair from the surgical area by clipping, shaving, or other mechanical means. Wash the area with an antiseptic solution. Ensure that hair does not come in contact with the implants or the wound surfaces. The surgical technique may profoundly influence the result of any implantation procedure. The surgery shall be carried out under aseptic conditions and in a manner that minimizes trauma at the implant site. After surgery close the wound, using either wound clips or sutures, taking precautions to maintain aseptic conditions.

Post-operative Assessment Observe each animal at appropriate intervals during the test period and record any abnormal findings, including local, systemic, and behavioral abnormalities.

Euthanasia At the termination of the experimental period, euthanize the animals with an overdose of anesthetic or by some other acceptable humane method (see ISO 10993-2).

Evaluation of Biological Response Evaluate the biological response by grading and documenting the macroscopic and histopathological test responses as a function of time. Compare the responses to the test material and control material. Carry out comparison of the control and the test implants at equivalent locations relative to each implant so that the effect of relative motion between tissue and implant is at a minimum. For a cylindrical specimen the region is midway between ends. With grooved cylindrical implants the center portions between the grooves as well as the flat top end surfaces of the implant are suitable for evaluation. For a non-solid or particulate material incorporated into a tube, the area at the end of the tube is the only available area for evaluation.

12

Macroscopic Assessment

×

Examine each implant site with the aid of a low magnification lens. Record the nature and extent of any tissue reaction observed.

× × ×

Preparation for Histology—Implant Retrieval and Specimen Preparation

Table 10.3 Selection of Test Periods for Long-Term Implantation in Subcutaneous Tissue, Muscle and Bond Implantation period (weeks) Species

12

26

52

78

104

Rats Guinea pigs Rabbits Dogs Sheep Goats Pigs

× × × × × × ×

× × × × × × ×

× × × × × × ×

× × × × ×

× × × ×

Excise the implant together with sufficient unaffected surrounding tissue to enable evaluation of the local biological response. Process the excised tissue blocks containing test or control implants for histopathological and other studies as appropriate. When conventional techniques are used, the tissue envelope may be opened before or after exposure to a fixative and the condition of the implant surface and tissue bed shall be reported. However, with this technique the tissue layers closest to the implant are usually destroyed.

136

Safety Evaluation of Medical Devices

When the implant/tissue surface is to be studied, embedding of the intact tissue envelope with the implant in situ using hard plastics is preferred. Appropriate sectioning or grinding techniques are employed for the preparation of histological sections. It shall be demonstrated that the technique of embedding in plastics does not markedly alter the interface tissue.

Histological Assessment The extent of response may be determined by measurement of the distance from the implant/tissue interface to unaffected areas with the characteristics of normal tissue and of normal vascularity. Record the section orientation in relation to the implant dimensions. Record the implant orientation, number of sections, and cutting geometry. The biological response parameters which shall be assessed and recorded include: (a) extent of fibrosis/fibrous capsule and inflammation; (b) degeneration as determined by changes in tissue morphology; (c) number and distribution as a function of distance from the material/tissue interface of the inflammatory cell types, namely polymorphonuclear leukocytes, lymphocytes, plasma cells, eosinophils, macrophages, and multinucleated cells; (d) presence of necrosis as determined by nuclear debris and/or capillary wall breakdown; (e) other parameters such as material debris, fatty infiltrations, granuloma; (f) for porous implant materials, the quality and quantity of tissue ingrowth.

per observation period shall be recorded. All specimens shall be accounted for and considered as part of the test. The techniques for taking histological sections shall be described.

Evaluation Macroscopic observations shall include the observations made on implant as well as the macroscopic appearance of the tissue surrounding the implant. The report shall include the results obtained from each histological examination. The report shall include a comparative evaluation of the biological responses to test and control materials, as well as a descriptive narrative of the biological response.

TEST METHOD FOR IMPLANTATION IN SUBCUTANEOUS TISSUE Field of Application This test material is used for assessing the biological response of subcutaneous tissue to an implanted material. The study may be used to compare the effect of different surface textures or conditions of the same material, or to assess the effect of various treatments or medications or a material.

Principle Insertion of the implants in the subcutaneous tissue of test animals. The method compares the biological response to implants of test specimens with the biological response to implants of control specimens made of materials which are established in clinical use.

Test Specimens

Description of test and control materials, material condition, fabrication, surface condition, and the shape and size of implants. Remember to specify the rationale for selection of control material(s). The surface preparation of the specimens can affect the tissue reaction. Therefore, the preparation procedure should be noted in the report. Report cleaning, handling, and sterilization techniques employed. If not done in-house, this information should be supplied by the manufacturer before the investigation commences.

Common provisions for preparation of test and control specimens as previously described. Implant sizes are based on the size of the test animal. Specimens made of sheet material shall be 10–12 mm in diameter and from 0.3 to 1 mm in thickness. Note: The subcutaneous site, deep to the panniculus carnosus muscle, is particularly suitable for the evaluation of polymeric sheet material, and for the evaluation of polymeric sheet material. In an intramuscular site, sheet material may become folded, which makes it difficult to assess the effect of the material per se. Bulk materials shall be fabricated into specimens 1.5 mm in diameter and 5 mm in length, and have radiused ends. Grooved specimens shall be 4 mm in diameter and 7 mm in length (see annex B). Note: Tissue ingrowth into the grooves minimizes tissue irritation caused by interface motion. Non-solid specimens (including powders) shall be prepared in tubes 1.5 mm in diameter and 5 mm in length.

Animals and Implantation

Test Animals and Implant Sites

Report on origin, age, sex, and strain of animals. Report housing conditions, diet, and mass of animals during the study period. The health of the animals shall be evaluated during the study. All observations, including unexpected death, shall be reported. Report insertion techniques. Report number of implants inserted per animal, per site, and per observation period.

The implants shall be inserted in the dorsal subcutaneous tissue of adult mice, rats, guinea-pigs, or rabbits. Select one species among these. Use at least three animals and sufficient sites to yield 10 specimens for each material and implantation period.

Retrieval and Histological Procedure

Implantation Along Dorsal Midline

The report shall include a description of the retrieval technique. The number of implants retrieved per animal and

Make an incision of the skin and make one or more subcutaneous pockets by blunt dissection. The base of the pocket

In the case of bone, the interface between the tissue and the material is of special interest. Evaluate the area of bone contact and the amount of bone in the vicinity of the implant as well as the presence of intervening non-calcified tissues. Note the presence of bone resorption and bone formation.

Implant Specimens

Implantation Procedure Select one of the procedures described below.

Chapter 10:

shall be more than 10 mm from the line of incision. Place one implant in each pocket. The implants shall not be able to touch one another. Note: Alternatively, the implants may by delivered by a trocar to the desired site.

Implantation in Neck In mice, make a 10 mm long incision above the sacrum and prepare a subcutaneous tunnel by blunt dissection towards the neck. Push one implant (for design see annex B) through the tunnel to position it at the neck. In rats, insert one implant of each of the control and candidate materials separately on each side of the neck. The implants shall not be able to touch one another. At some distance from the implant, close the tunnel with stitches of appropriate suture material to prevent the implant from moving.

Implantation Period To ensure a steady state of biological tissue response the implantation period(s) shall be as specified in Tables 9.2 and 9.3.

Evaluation of Biological Response The evaluation shall take into account the items specified earlier.

Implantation Biology and Studies

137

of a material known to evoke a minimal tissue reaction, in a location opposite to the test materials.

Implantation Procedure Implantation shall be by hypodermic needle or trocar. For larger implants other appropriate surgical implantation techniques may be used. Implant test specimens into the body muscle with the long axis parallel to the muscle fibers. For rabbit paravertabral muscles, implant four specimens of the test materials along one side of the spine, 25–50 mm from the midline and parallel to the spinal column, and about 25 mm apart form each other. In similar fashion implant four specimens of the control material in the contralateral muscle of each animal.

Implantation Period To ensure a steady state of biological tissue response, the implantation period(s) shall be as specified in Tables 9.2 and Table 9.3.

TEST METHOD FOR IMPLANTATION IN BONE Field of Application

TEST METHOD FOR IMPLANTATION IN MUSCLE Field of Application

This test method is used for assessing the biological response of bone tissue to an implanted material. The study may be used to compare the effect of different surface textures or conditions of the same material, or to assess the effect of various treatments or modifications of a material.

This test is used for assessing the biological response of muscle tissue to an implanted material.

Principle

Principle Insertion of the implant in the muscle of a test animal. The method compares the biological response to implants of test specimens with the biological response to implants of control specimens made of materials which are established in clinical use.

Test Specimens Common provisions for preparation of test and control specimens are described earlier. Implant sizes are based on the size of the muscle group chosen. For rabbit paravertabral muscles, implants of a width of 1–3 mm with a length of approximately 10 mm shall be used. The specimens shall have rounded edges and the ends finished to a full radius.

Test Animals and Implant Sites Insert the implants in the muscle tissue of rabbits or other animals. Ensure that the muscles are of sufficient size to accommodate the implant specimens. Use only one species per test. Note: The paravertabral muscles of rabbits are the preferred implant sites. Alternatively, the gluteal muscles of rats, or the thigh muscles of rabbits may be used. Use at least three animals and sufficient implant sites to yield eight test specimens and eight control specimens for each implantation period. In cases where the control material is expected to elicit more than a minimal response, use two specimens of this control. Implant two additional control specimens, composed

Insertion of the implants into the bone tissue of test animals. The method compares the biological response to implants of test specimens with the biological response to implants of control specimens made of materials which are established in clinical use.

Shape of Implant Specimens The specimens may be screw-shaped or threaded to provide initial stability of the implants in bone. If preparation of a screw shape is impractical, a cylinder shape may be used.

Size of Test Specimens Implant sizes are based on the size of the test animal and bone chosen. The following dimensions shall be considered: (a) rabbits: cylindrical implants 2 mm in diameter and 6 mm in length; (b) dogs, sheep, and goats: cylindrical implants 4 mm in diameter and 12 mm in length; (c) rabbits, dogs, sheep, goats, and pigs: 2 mm to 4.5 mm orthopedic bone screw-type implants.

Test Animals The implants shall be inserted into the bone of dogs, sheep, goats, pigs, or rabbits. Select one species among these. Species differences are important in bone physiology, and should be assessed before implantation procedures are initiated. At least four rabbits, or at least two each of other animals, shall be used for each implantation period.

Implant Sites Equivalent anatomical sites shall be used for test and control specimens. The test implants shall be contralateral to the

138

Safety Evaluation of Medical Devices

control implants. Select the implant site to minimize the risk of mobility of the implant. Note: The femur and tibia are suitable. Other sites may be considered. The number of implant sites shall be as follows;

Polymeric and Ceramic Control Materials

(a) in each rabbit there shall be a maximum of six implant sites: three for test specimens and three for control specimens; (b) in each dog, sheep, goat, or pig, there shall be a maximum of 12 implant sites; six for test specimens and six for control specimens. Do not insert more than 12 specimens in any one animal.

Implantation studies can also be conducted to evaluate the longer term (subchronic and chronic) potential for devices to elicit systemic toxicity, or to evaluate the carcinogenic potential of devices. Uses for these cases are addressed in Chapters 12 and 13.

The size, mass, and age of the animal and the implant site chosen should ensure that the implant placement does not cause significant risk of pathological fracture of the test site. In younger animals it is especially important to ensure that the implants avoid the epiphyseal area or other immature bone.

Implantation Procedure Perform bone preparation using low drilling speed and intermittent drilling with profuse irrigation with physiological saline solution and suction, because overheating will result in local tissue necrosis. It is important that the diameter of the implant and the implant bed in the bone match well enough to avoid ingrowth of fibrous tissue. Expose the cortex of each femur or tibia and drill the appropriate number of holes to receive implants. For rabbits, prepare up to three holes; for larger animals prepare up to six holes. Ream to final diameter or tap screw thread before insertion. Insert cylinders by finger pressure to allow press fit. Tighten screw-shaped implants in place with an instrument capable of delivering a predetermined torque. Record the torque.

Information on non-metallic control materials is to be found in ASTM F 748, 763, and 981.

Implantation as a Method for Other Endpoints

LONG-TERM IMPLANT STUDIES Chapter 12 will address the issues and considerations involved in evaluating the systemic effects of long-term implant devices, and Chapter 13 the case of evaluation of materials for potential carcinogenicity. There are also local tissue and body/implant interactions that must be evaluated (Leninger et al., 1964). The spectrum of interactions can be thought of as presented in Table 10.4. Such interactions are assessed in long-term studies which may or may not include the eventual retrieval of the implant itself from the host. Retrieval studies seek to study the biological and device-related performance characteristics under actual conditions of use, and to determine the efficacy, reliability, and biocompatibility (safety) of medical devices. As such, retrieval studies have seven objects: 1. 2. 3. 4. 5. 6. 7.

Enhanced patient management Recognition of complications Device design criteria Evaluation of patient/prosthesis matching Elimination of complications Identification of interactions Elucidation of mechanisms of interactions.

Implantation Period To ensure a steady state of biological tissue response the implantation period(s) shall be as specified in Tables 10.2 and Table 10.3.

CONTROL MATERIALS Response The biological response to these materials is not defined as to response, but rather the response is used as a reference against which a reaction to another material is compared. As a porous control material is not available at present, it is acceptable to use a dense control material for comparative purposes. If the most appropriate control material is expected to elicit a tissue response greater than that normally observed with the control materials cited in this annex, samples of these latter materials may be implanted as controls to check the surgical technique.

Table 10.4

Host/Implant Interactions

Effects of implant on host

Effects of host on implant

Local Blood material interactions Protein adsorption Coagulation Fibrinolysis Platelet reactions Complement activation Blood Leukocyte reactions Hemolysis Toxicity Derangements of healing Encapsulation Foreign body reaction Pannus formation Infection Tumorigenesis

Physical Abrasive wear Fatigue Stress-corrosion Degeneration Dissolution

Metallic Control Materials Stainless steel, cobalt-chromium, titanium, and titanium alloys are used to fabricate control specimens. The biological response to these materials has been well characterized by their extensive use in research and clinical practice. See for further information ISO 5832, Parts 1–8.

Systemic Embolization Thrombus Hypersensitivity Alteration of lymphatic system

Biological Adsorption of tissue Substances Enzymatic degradation Calcification

Chapter 10:

Implantation Biology and Studies

Implants can fail for any of six different categories of causes:

Table 10.5

1. 2. 3. 4. 5. 6.

Necropsy periods (weeks after insertion of implants)

Thrombosis and thromboembolism Device-assisted infection Inappropriate healing Degradation, fracture Adverse local tissue reaction Adverse systemic reaction.

ASTM standard practice F981-87 (“Standard Practice for Assessment of Compatibility of Biomaterials [Nonporous] for Surgical Implants with Respect to Effect of Materials on Muscle and Bone”) provides a framework for evaluating long-term host/implant interactions. The practice provides a series of experimental protocols for biological assays of tissue reaction to nonporous, nonabsorbable biomaterials for surgical implants. It assesses the effects of the material on animal tissue in which it is implanted. The specified experimental protocol is not designed to provide a comprehensive assessment of the systemic toxicity, carcinogenicity, teratogenicity, or mutagenicity of the material. It applies only to materials with projected applications in human subjects where the materials will reside in bone or soft tissue in excess of 30 days and will remain unabsorbed. Applications in other organ systems or tissues may be inappropriate and are therefore excluded. Control materials will consist of any one of the metal alloys in ASRM Specifications F67, F75, F90, F136, F138, or F562, or ultra-high-molecular-weight polyethylene as stated in ASTM Specifications F648 or USP polyethylene negative control. Referenced ASTM Standards include: F67 F75 F86 F90 F136 F138 F361 F469 F562 F648 F673

Specification for Unalloyed Titanium for Surgical Implant Applicationsa Specification for Cast Cobalt–Chromium–Molybdenum Alloy for Surgical Implant Applicationsb Practice for Surface Preparation and Marking of Metallic Surgical Implantsa Specification for Wrought Cobalt–Chromium–Tungsten– Nickel Alloy for Surgical Implant Applicationsb Specification for Wrought Titanium 7A1-4V ELI Alloy for Surgical Implant Applicationsa Specification for Stainless Steel Bars and Wire for Surgical Implants (Special Quality)a Practice for Assessment of Compatibility of Metallic Materials for Surgical Implants with Respect to Effect of Materials on Tissueb Practice for Assessment of Compatibility of Nonporous Polymeric Materials for Surgical Implants with Regard to Effect of Materials on Tissuec Specification for Wrought Cobalt–Nickel–Chromium– Molybdenum Alloys for Surgical Implant Applicationa Specification for Ultra-High-Molecular-Weight Polyethylene Powder and Fabricated Form for Surgical Implantsa Practice for Short-Term Screening of Implant Materialsa.

The practice describes the preparation of implants; the number of implants and test hosts, test sites, exposure schedule, implant sterilization techniques; and methods of

a

Annual Book of ASTM Standards, Vol. 13.01. b Discontinued. See 1986 Annual Book of ASTM Standards Vol. 13.01. c Discontinued. See 1987 Annual Book of ASTM Standards Vol. 13.01.

139

Intervals of Sacrifice Number of animals to be necropsied

12 weeks 26 weeks 52 weeks 104 weeks

Rat

Rabbit

Dog

4 4 4 —

4 4 4 —

2 2 2 2

implant retrieval and tissue examination of each test site. Histological criteria for evaluating tissue reaction are provided. A test protocol for comparing the local tissue response evoked by biomaterials is specified, from which medical implantable devices might ultimately be fabricated, with the local tissue response elicited by control materials currently accepted for the fabrication of surgical devices. Currently accepted materials are the metals, metal alloys, and polyethylene previously specified which are standardized on the basis of acceptable long-term clinical experience. The controls consistently produce cellular reaction and scar to a degree that has been found to be acceptable to the host. Rats (acceptable strains such as Fischer 344), New Zealand rabbits, and dogs may be used as test hosts for soft tissue implant response. It is suggested that the rats be age and sex matched. Rabbits and dogs may be used as test hosts for bone implants. The sacro-spinalis, paralumbar, gluteal muscles and the femur or tibia can serve as the test site for implants. However, the same site must be used for test and material implants in all the animal species. Table 10.5 contains a suggested minimum number of study animals and a suggested schedule for the necropsy of animals. Each implant shall be made in a cylindrical shape with hemispherical ends (see below for sizes). If the ends are not hemispherical, this must be reported. Each implant shall be fabricated and finished and its surface cleaned in a manner appropriate for its projected application in human subjects, in accordance with ASTM Practice F86. Reference metallic specimens shall be fabricated from materials such as the metal alloys in ASTM Specifications F67, F75, F90, F138 or F562, or polymeric polyethylene USP negative control plastic. Suggested sizes and shapes of implants for insertion in muscle are as follows: For rats 1-mm diameter by 2-cm long cylindrical implants. For rabbits 2-mm diameter by 10–15-mm long cylindrical implants. For dogs 6-mm diameter by 18-mm long cylindrical implants. If fabrication problems prevent preparing specimens 1 mm in diameter, alternative specimen sizes are 2-mm diameter by 6-mm long for rats and 4-mm diameter by 12-mm long for rabbits. If these alternate dimensions are used, such should be reported and such use justified. Sizes and shapes of implants for insertion in bone are as follows: For rabbits 2-mm diameter by 6-mm long cylindrical implants. For dogs 4-mm diameter by 12-mm long cylindrical implants.

140

Safety Evaluation of Medical Devices

If the length of the bone implants needs to be less than that designated because of anatomical constraints, such should be reported.

Number of Test and Control Implants In each rat, due to size, there should be two implants: one each for test and control material implant. In each rabbit, due to size, there should be six implants: four for test materials and two control material implants. In each dog, there should be twelve implants: eight for test materials and four control material implants.

Conditioning Remove all surface contaminants with appropriate solvents and rinse all test and control implants in distilled water prior to sterilization. It is recommended that the implant materials be processed and cleaned in the same way the final product will be. That is, clean, package, and sterilize all implants in the same way as used for human implantation. After final preparation and sterilization, handle the test and control implants with great care to ensure that they are not scratched, damaged, or contaminated in any way prior to insertion. Report all details of conditioning.

Implantation Period Insert all implants into each animal at the same surgical session so that implantation periods run concurrently. The implantation period is 52 weeks for rats and rabbits and 104 weeks for dogs, with interim sacrifices at 12, 26, and 52 weeks (see Table 10.5).

Implantation (Muscle) Place material implants in the paravertebral muscles of the adult rats, rabbits, or dogs in such a manner that they are directly in contact with muscle tissue. Introduce material implants in dogs by the technique of making an implantation site in the muscle by using a hemostat to separate the muscle fibers. Then insert the implant using plastic-tipped forceps or any tool that is nonabrasive to avoid damage to the implant. Do not insert more than twelve implant materials in each dog. Introduce material implants in rabbits and rats using a sterile technique. Sterile disposable Luer-lock needles may be used to implant the material implants into the paravertebral muscles along the spine. In rats insert a negative control implant on one side of the spine and a test material implant on the other side. In rabbits implant one negative control material on each side of the spine and implant two test materials on each side of the spine. If larger diameter specimens are used, an alternative implantation technique such as that described above should be employed.

Implantation (Femur) Expose the lateral cortex of each rabbit femur and drill three holes 1/16 in. (1.6 mm) through the lateral cortex using the technique and instrument appropriate for the procedure. For dogs, make the holes 1/8 in. (3.2 mm) in diameter; make six holes in each femur. Into each one of these holes, insert one of the implants by finger pressure. Then close the wound. Caution should be taken to minimize the motion of the implant in the tissue on the desired result.

Postoperative Care Care for the animals in accordance with accepted standards as outlined in the Guide for the Care and Use of Laboratory Animals. Carefully observe each animal during the period of assay and report any abnormal findings. Infection or injury of the test implant site may invalidate the results. The decision to replace the animal so that the total number of retrieved implants will be as represented in the schedule should be dependent upon the design of the study. If an animal dies prior to the expected date of sacrifice, necropsy it to determine the cause of death. Replacement of the animal to the study should be dependent upon the design of the study. Include the animal in the assay of data if the cause of death is related to the procedure or test material.

Sacrifice and Implant Retrieval Euthanize animals by a humane method at the intervals listed in Table 9.5. The necropsy periods start at 12 weeks because it is assumed that acceptable implant data have been received for earlier periods, such as 1, 4, and 8 weeks from short-term implant testing. At necropsy, record any gross abnormalities of color or consistency observed in the tissue surrounding the implant. Remove each implant with an intact envelope of surrounding tissue. Include in the tissue sample a minimum of a 4-mm thick layer of tissue surrounding the implant. If less than a 4-mm thick layer is removed, report such as the case.

Postmortem Observations Necropsy all animals that are sacrificed for the purposes of the assay or die during the assay period in accordance with standard laboratory practice. Establish the status of the health of the experimental animals during the period of the assay.

Histological Procedure Tissue Sample Preparation Prepare two blocks from each implantation site. Process the excised tissue block containing either a test implant or control implant for histopathological examination and such other studies as are appropriate. Cut the sample midway from end to end into appropriate size for each study. Record the gross appearance of the implant and the tissue. If special stains are deemed necessary, prepare additional tissue blocks or slides, or both, and make appropriate observations.

Histopathological Observations Compare the amount of tissue reaction adjacent to the test implant to that adjacent to a similar location on the control implant with respect to thickness of scar, presence of inflammatory or other cell types, presence of particles, and such other indications of interaction of tissue and material as might occur with the actual material under test (Pizzoferrato et al., 1988; Rahn et al., 1982). A suggested method for the evaluation of tissue response after implantation can be found in Turner et al. (1973), as summarized in Table 10.6. Pathologists may choose to use this scoring system while comparing the negative control to the test material as an aid in their evaluation. The overall toxicity of the test material as compared to the negative control is to be evaluated independently for all time periods. Table 10.7 provides a suggested format for evaluation and scoring.

Chapter 10:

Table 10.6

Suggested Methods for Tissue Response Evaluation

Number of elements*

Score

0 1–5 6–15 16–25 26 or more

0 0.5 1 2 3

Degree of necrosis Not present Minimal present Mild degree of involvement Moderate degree of involvement Marked degree of involvement

0 0.5 1 2 3

Overall toxicity rating of test samples Nontoxic Very slight toxic reaction Mild toxic reaction Moderate toxic reaction Marked toxic reaction

0 1 2 3 4

*Cellular elements to be evaluated based upon the number of elements in high power field (470×), average of five fields.

Report The final report shall include the following information: All details of implant characterization, fabrication, conditioning (including cleaning, handling, and sterilization techniques employed) Procedures for implantation and implant retrieval Details of any special procedure (such as unusual or unique diet fed to test animals) The observations of each control and test implant as well as the gross appearance of the surrounding tissue in which the implants were implanted The observation of each histopathological examination and the pathologist’s evaluation as to toxicity of test material provided. The ASTM standard practice is based on the research techniques utilized by Cohen (1959) and by Laing et al. (1967) Table 10.7

Suggested Evaluation Format and Scoring Range

Animal number Duration of implant (weeks) Sample description Gross response Histopath (number) Score Necrosis Degeneration Inflammation Polymorphonuclear leukocytes Lymphocytes Eosinophils Plasma cells Macrophages Fibrosis Giant cells Foreign body debris Fatty infiltration Relative size of involved area (mm) Histopathologic toxicity rating

__________________________ __________________________ __________________________ __________________________ __________________________ 0.5

1

2

3

4

__________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________ __________________________

Implantation Biology and Studies

141

in the early 1960s. These studies involved the implantation of metal cylinders in the paravertebral muscle of rabbits. The biological reaction to the cylinders was described as the thickness of the fibrous membrane or capsule formed adjacent to the implant. The thickness of the capsule and the presence of inflammatory cells were used as a measure of the degree of adverse reaction to the test material. As first published in 1972, Practice F361 was a test for the biological response to metallic materials. The scope had been expanded beyond that of the published reports to include bone as well as muscle as an implant test site. To avoid species-specific reactions, the method called for the use of rats and dogs as well as rabbits. Cylindrical test specimens with rounded ends were used to avoid biological reactions associated with sharp corners or other variations in specimen shape. In 1978, Practice F469 was published as a parallel document for the test of polymeric materials. In that the methods are essentially the same, the scope of F361 has been expanded to include the testing of specimens made of metallic, polymeric, or ceramic materials, thereby including and superseding F469. Porous or porous-coated materials are specifically excluded since the response to such materials includes ingrowth of tissue into the pores. As a result, the method of tissue fixation and sectioning and the evaluation scheme are substantially different. Stainless steel, cobalt-chromium, and titanium alloys are used as reference materials since the biological response to these materials has been well characterized by their extensive use in research. The response to these materials is not defined as compatible, but rather the response is used as a reference against which reactions to other materials is compared. This practice is a modification of the original Practice F361 in that it only involved long-term test periods. The short-term response to materials is to be evaluated using Practice F763. Special methods exist to reduce the impact of relative motion at the implant/tissue interface (Geret et al., 1980a, b). This practice was revised in 1987 to allow for alternative specimen dimensions for rats and rabbits for muscle implantation. The original specimen dimensions were intended to be implanted through a needle, which was a change from F361 and F469. The alternate dimensions restore those specified since 1972 which some members felt were more appropriate for some material types.

CONSIDERATIONS One problem of the implantation tests is a tendency for the strips or prototype implant devices to migrate from their implantation sites, even to subcutaneous positions, and this often prolongs the search for them. Nevertheless, the test is an effective detection system for toxic ingredients of solid materials which leach in contact with tissue fluid. It is important to recognize the microscopic effects of the standard negative control strips (such as the USP additive-free polyethylene). These reactions are typical of skeletal muscle in contact for a week or more with a foreign body and comprise mild mononuclear cell infiltration, multinucleated giant cell formation, fibroplasia, slight dystrophic calcification, muscle fiber atrophy, and centripetal migration of sarcolemmal nuclei. Traumatic hemorrhage is also common. Positive reactions are

142

Safety Evaluation of Medical Devices

similar but more pronounced and additionally include focal necrosis and exudation, particularly of heterophils. These are useful tests, not only for finished products, but also to identify unacceptable changes in formulation or manufacturing processes such as the introduction of chlorinating cycles to remove bloom on latex catheters. For materials to be utilized in long-term implants, these tests are the only means of accurately predicting long-term tissue and systemic interactions. The finished state and handling of all components, but particularly metals, in an implant are critical variables, as the primary tissue/implant interactions revolve around surface effects such as the ionization of metals (Ferguson et al., 1960). Such interactions are so predominantly surface interactions (Kordan, 1967) that the surface conditions of an implant, such as porosity and pore size, are critical (Goldhaber, 1961, 1962). The actual site of implantation in the body also influences the nature of interaction significantly (Kaminski et al., 1968).

REFERENCES ASTM F 748, Practice for Selecting Generic Biological Test Methods for Materials and Devices. ASTM F 763, Practice for Short-Term Screening of Implant Materials. ASTM F 981, Practice for Assessment of Compatibility of Biomaterials for Surgical Implants with Respect to Effect of Materials on Muscle and Bone. Black, J. (2000). Biological Performance of Materials, 4th Edn. Taylor & Francis, New York. British Standard (1981a). No. 5736/2, 5736/3 (1981b), 5736/4 (1981c), 5736/5 (1982). British Standards Institution, London. Cohen, J. (1959). Assay of foreign-body reaction. J. Bone Joint Surg., 41A: 152–166. Ferguson, Jr., A.B., Laing, P.G., and Hodge, E.S. (1960). The ionization of metal implants in living tissues. J. Bone Joint Surg., 42A: 77–90. Gad, S.C. (2006). Animal Models in Toxicology, 2nd Edn. Taylor and Francis, Philadelphia. Greco, R.S. (1994). Implantation Biology. CRC Press, Boca Raton, FL. Geret, V., Rahan, B.A., Mathys, R., Straumann, F., and Perren, S.M. (1980a). In vivo testing of tissue tolerance of implant materials: Improved quantitative evaluation through reduction of relative motion at the implant tissue interface. In: H.K. Uhthoff (Ed.), Current Concepts of Internal Fixation of Fracture. Springer Verlag, New York.

Geret, V., Rahan, B.A., Mathys, R., Straumann, F., and Perren, S.M. (1980b). A method for testing tissue tolerance for improved quantitative evaluation through reduction of relative motion at the implant-tissue interface. In: G.D. Winter, J.L. Leray and K. de Groot (Eds.), Evaluation of Biomaterials. John Wiley & Sons, Ltd, London. Goldhaber, P. (1961). The influence of pore size on carcinogenicity of subcutaneiously implanted millipore filters. Proc. Am. Assoc. Cancer Res., 3: 228. Goldhaber, P. (1962). Further observations concerning the carcinogenicity of millipore filters. Proc. Am. Assoc. Cancer Res., 3: 323. Greco, R.S. (1994). Implantation Biology. CRC Press, Boca Raton, FL. Guelcher, S.A. and Hollinger, J.O. (2006). An Introduction to Biomaterial. Taylor & Francis, Boca Raton, FL. ISO 10993 (1995). Part 6: Tests for Local Effects after Implantation. Kaminski, E.J., Oblesby, R.J., Wood, N.K., and Sandrik, J. (1968). The behavior of biological materials at different sites of implantation. J. Biomed. Mater. Res., 2: 81–88. Kordan, H.A. (1967). Localized interfacial forces resulting from implanted plastics as possible physical factors involved in tumor formation. J. Theroret. Biol., 17: 1–11. Laing, P.G., Ferguson, Jr., A.B., and Hodge, E.S. (1967). Tissue reaction in rabbit muscle exposed to metallic implants. J. Biomed. Mater. Res., 1: 135–149. Leininger, R.I., Mirkovitch, V., Peters. A. et al. (1964) Change in properties of plastics during implantation. Trans. Amer. Soc. Artif. Int. Organs, 10: 320–321. Pizzoferrato, S., Savarino, L., Stea, S. and Tarabusi, C. (1988). Result of histological grading on 100 cases of hip prosthesis failure. Biomaterials, 9: 314–318. Rahn, B.A., Geret, V., Capaul, C., Lardi, M., and Solothurnmann, B. (1982). Morphometric evaluation of tissue reaction to implants suing low cost digitizing techniques. In: A.J.C. Lee, T. Albrektsson and P.I. Branemark (Eds.), Clinical Application of Biomaterials. John Wiley & Sons Ltd., New York. Raner, B.D., Hoffman, A.S., Scholen, F.J., and Temons, J.E. (2004). Biomaterials Science, 2nd Edn. Elsevier, New York. Turner, J.E., Lawrence, W.H., and Autian, J. (1973). Subacute toxicity testing of biomaterials using histopathologic evaluation of rabbit muscle tissue, J. Biomed. Mater. Res., 7: 39–58. USP (1996). United States Pharmacopoeia XXIII. US Pharmacopoeia Convention, Washington, D.C. USP (2007). United States Pharmacopoeia 31 and NF26, U.S Pharmacopoeia Convention, Washington, DC. von Recum, A.R. (1998). Handbook of Biomaterials Evaluation. Taylor & Francis, Philadelphia, PA.

11 Acute Systemic Toxicity Testing Device Safety Evaluation

screens, and are normally the type of study done to satisfy regulatory requirements or to provide a more thorough early characterization or prediction of toxicity. In pharmaceutical development, rarely would an acute test be sufficient to support registration, but it may be required as part of an overall package. These protocols may resemble range-finding tests, but they call for collection of more data. A list of the types of data that can be obtained in well-conducted acute toxicity tests is given in Table 11.1. Given that these studies usually include control groups, the classical or traditional design is the most common because it allows for the most straightforward statistical analyses. In addition, while the use of staggered dosing days for different groups is still a fairly common practice, data analyses may be more sensitive if all animals are dosed on the same day, requiring that one has preliminary range-finder data that permit selection of appropriate dosages. Studies of more than one species and/or more than one route should be limited to those instances where they are required by statute. In general, traditionally designed acute toxicity tests can be divided into three types that can be called the minimal acute toxicity test, the complete acute toxicity test, and the supplemented acute toxicity test. Of these, the minimal protocol is by far the most common and is discussed first. The other two represent increasing orders of complexity as additional parameters of measurement are added to the basic minimal study.

INTRODUCTION What is acute toxicity testing and why does it warrant a separate chapter in this book? Of what value is it in ensuring the safety of medical device or biomaterial? Acute toxicity testing is the defining and evaluation of the toxic syndrome (if any) produced by a single dosage of (or, in the case of an implanted device, the initial exposure of the body to) a device. Historically, the main focus of these tests has been lethality determinations and the identification of overt signs and symptoms of overdosage. For a complete historical perspective, see Rhodes (2000), Gad and Chengelis (1999), Auletta (1998), or Piegorsh (1989). A more enlightened and modern view holds that, especially for pharmaceutical agents, lethality in animals is a relatively poor predictor of hazard in man (Gad and Chengelis, 1999). The current trend is toward gaining increasing amounts of more sophisticated data from these tests, such as in the expanded acute studies done to enable so called “Phase 0” clinical trials. The various types of acute study designs, their utility in pharmaceutical product testing, and the resultant sample data are discussed in this chapter. In the pharmaceutical industry, acute toxicity testing has uses other than for product safety determinations. First, as in other industries, acute toxicity determinations are part of industrial hygiene or occupational health environmental impact assessments. These requirements demand testing not only for finished products but frequently of intermediates as well. These issues and requirements, however, are discussed in Chapter 2 and are not directly addressed here. For medical devices, acute systemic toxicity typically means the ISO-10993 or pharmacopeail (such as USP and other pharmacopeias) “acute systemic toxicity” testing is actually lethality testing using one or more extracants (“eluants”) from the material of interest. It is designed to screen broadly for the presence of any extractable materials which may be lethal to the intact organism. The “model” employed is the albino mouse (17–23 g in weight). USP section <88> provides guidance on preparation of materials and test conduct. Following is an example protocol: Extractions media area: ● ● ● ● ●

Clinical Signs The nonlethal parameters of acute toxicity testing have been extensively reviewed by Sperling (1976) and Balazs (1970, 1976). Clinical observations or signs of toxicity are perhaps the most important aspect of a minimal acute toxicity test because they are the first indicators of drug- or chemicalrelated toxicity or morbidity, and they are necessary in the interpretation of other data collected. For example, body weight loss (or a reduction in body weight gain) would be expected if an animal had profound CNS depression lasting several hours. With regard to clinical signs and observations, there are some basic definitions that should be kept in mind. Symptomatology is the overall manifestation of toxicity. Signs are overt and observable events (Brown, 1983). Symptoms are the subjective impressions of a human patient (e.g., headache) and cannot be described or reported by speechless animals (Balazs, 1970). Clinical signs can be reversible or irreversible. Reversible signs are those that dissipate as the chemical is cleared from the body or tolerance develops (Chan et al., 1982) and are generally not accompanied by permanent organ damage. Irreversible signs are those that do not dissipate and are generally accompanied by organic damage. Signs can also represent a normal biological or pharmacological response (Chan et al., 1982).

Sodium chloride 0.9% (physiological saline solution) Polyethylene glycol 400 5% ethanol in 0.9% aqueous saline Vegetable (cottonseed or sesame) oil If not covered by any of the above, device product vehicle should be employed.

ACUTE SYSTEMIC TOXICITY CHARACTERIZATION Acute systemic toxicity studies are performed to more completely define the acute toxicity of a drug. They are more extensive and time consuming than range-finding tests or 143

144

Safety Evaluation of Medical Devices

STUDY DAY Prepare Eluates

0

1

2

3

Weigh Animals Weigh and Dose 5 Test Mice 5 Control Mice

Figure 11.1

OBSERVE FOR MORTALITY

Extract or Blank

Dose (/kg)

Route

Injection Rate (µL/second)

SAL

50 mL

IV

100

5% ETOH

50 mL

IV

100

PEG 400

10 g

IP

NA

CSO

50 mL

IP

NA

USP acute systemic toxicity test.

Acclimatization period

Observation period Death (1)

−14

−10

−6

−2

1

3

Animals chosen, dose (1000 mg/kg) 5/sex

5

7

9 11 (Days)

Termination

Results: 1/10 deaths • LD50 > 1000 mg/kg ••

Figure 11.2

Example of typical limit test protocol.

−7∗ one or two species

0

2∗∗

14∗∗∗

Dose

Figure 11.3 Exploratory IND enabling acute study. *One species only (selections justified with in vitro data) or two species. **Clinical Chemistry and Hematology, Gross Necropsy and Histopathology, Interim Groups and Histopathology. ***Clinical Chemistry and Hematology, Gross Necropsy and Histopathology, In Terminal Groups. Also body weights (days 1, 2, 4, 7 and 14). Clinical observations (daily). Doses, 100× what you want to do in clinic (microdoses).

Chapter 11:

Table 11.1 Information, Including Lethality, That Can Be Gained in Acute Toxicity Testing Lethality/Mortality LD50 with confidence limits Shape and slope of lethality curves Estimation of maximum nonlethal dose or minimum lethal dose (LD01) Time to dose estimates Clinical signs Times of onset and recovery Thresholds Agonal vs. nonagonal (i.e., do signs occur only in animals that die?) Specific vs. general responses Separation of dose-response curves from lethality curves Body weight changes Actual loss vs. decreased gain Recovery Accompanied by changes in feed consumption Changes in animals that die vs. those that survive Target organ identification Gross examinations Histological examinations Clinical chemical changes Hematological changes Specialized function tests Immunocompetency Neuromuscular screening Behavioral screening Pharmacokinetic considerations Different routes of administration yielding differences in toxicity Plasma levels of test article Areas under the curves, volume of distribution, half-life Metablic pattern of test article Distribution to key organs Relationship between plasma levels and occurrence of clinical signs

Table 11.2

Acute Systemic Toxicity Testing Device Safety Evaluation

145

For example, an antidepressant would be expected to cause decreased activity and some ataxia. These symptoms are generally reversible and can lead to secondary, nonspecific signs—nonspecific in that any number of agents or stimuli can evoke the same response, and secondary in that they are probably not due (at least, one has no evidence to determine otherwise) to the direct action of the test article. Responses can also be abnormal, in that they are not due to a homeostatic process. The increases in serum urea and creatinine due to kidney damage, for example, are abnormal responses. These are often irreversible, but this is not always the case, depending on the repair capacity or functional reserves of the target organ. These abnormal responses may also be called primary effects because they reflect the direct action of a test article. Agonal signs are those occurring immediately prior to, or concomitantly with, death. They are obviously irreversible, but not necessarily reflective of a specific effect of a test article. For example, regardless of the cause, labored breathing will occur in a moribund animal. It is, therefore, important to distinguish between signs that occur in animals that die and those that do not. It should also be kept in mind that agonal signs may mask (make it difficult or impossible) to observe other signs, including those clearly seen at lower doses. In their simplest form, clinical observations are those done on an animal in its cage, or, preferably, in an open plane, such as on the top of a counter or laboratory cart. These are considered passive observations. One can gain even more information by active examination of the animal, such as the animal’s response to stimulation. Fowler and Rutty (1983) divide their clinical evaluation of toxicity into those signs scored by simple observations (e.g., ataxia), those scored by provocation (e.g., righting reflex), those scored in the hand (e.g., mydriasis) and those scored by monitoring (e.g., rectal temperature). Cage pans should always be examined for unusually large or small amounts of excreta, or excreta of abnormal color or consistency. A list of typical observations is summarized in Table 11.2. A more extensive

Clinical Observation in Acute Toxicity Tests

Organ system

Observation and examination

Common signs of toxicity

CNS and somatomotor

Behavior Movements Reactivity to various stimuli Cerebral and spinal reflexes Muscle tone Pupil size Nostrils Character and rate Palpation of cardiac region Events Abdominal shape Feces consistency and color Vulva, mammary glands Penis Perineal region Color, turgor, integrity Conjunctiva, mouth Eyelids Eyeball Transparency Rectal or paw skin temperature Injection site General condition

Unusual aggressiveness, unusual vocalization, restlessness, sedation Twitch, tremor, ataxia, catatonia, paralysis, convulsion Irritability, passivity, anesthesia, hyperesthesia Sluggishness, absence of reflex Rigidity, flaccidity Miosis, mydriasis Discharge (color vs. uncolored) Bradypnea, dyspnea, Cheyne-Stokes breathing, Kussmaul breathing Thrill, bradycardia, arrhythmia, stronger or weaker beat Diarrhea, constipation Flatulence, contraction Unformed, black or clay colored Swelling Prolapse Soiled Reddening, flaccid skinfold, eruptions, piloerection Discharge, congestion, hemorrhage, cyanosis, jaundice Ptosis Exophthalmos, nystagmus Opacities Subnormal, increased Swelling Abnormal posture, emaciation

Autonomic nervous system Respiratory Cardiovascular Gastrointestinal

Genitourinary

Skin and fur Mucous membranes Eye

Others

Source: Balazs (1970).

146

Safety Evaluation of Medical Devices

table has been prepared by Chan et al. (1982). Given the fact that the number of different signs displayed is not infinite and that some signs are simply easier to discern than others, most clinical signs are referable to the CNS (e.g., lack of activity), the GI tract (e.g., diarrhea), or the general autonomic nervous system (e.g., increased salivation or lacrimation). Other signs can be detected by a well-trained observer, but are, nonetheless, less common than those described above. Respiratory distress can be diagnosed by examining the animal’s breathing motions and listening for breathing noises. Cardiovascular signs are generally limited to pallor, cyanosis, and/or hypothermia. Changes in cardiac function can be difficult to detect in small animals, and generally consist of “weak” or “slow” breathing. Arrhythmias can be difficult to detect because the normal heart rate in a rodent is quite rapid. Given the subjective nature of recognizing clinical signs, careful steps must be taken to ensure uniformity (is the animal depressed or prostrated?) of observation so that the data can be analyzed in a meaningful fashion. There are three ways of achieving this. First, signs should be restricted to a predefined list of simple descriptive terms, such as those listed in Table 5.4 or in Appendix B.

Second, if a computerized data acquisition system is unavailable, the use of standardized forms will add uniformity to the observation and recording processes. An example of such a form is shown in Figure 11.4. Third, technicians should be trained in studies (not intended for regulatory submission) using material of known toxicity, so that all personnel involved in such evaluations are using the same terminology to describe the same signs. Animals should be observed continuously for several hours following dosing. Times of observation should be recorded as well as the actual observations. After the first day of the study, observations generally need only to consist of brief checks for sign remission and the development of new signs of morbidity. Data should be collected in such a way that the following could be concluded for each sign: (1) estimated times of onset and recovery, (2) the range of threshold dosages, and (3) whether signs are directly related (primary) to the test article. Incidences are broken down by dosage group and sex. These data illustrate the fact that mortality can censor (preclude) the occurrence of clinical signs. Note that reduced fecal output was a more frequent observation at the intermediate dosages because most of the animals died at the higher dosages.

ACUTE OBSERVATION RECORD (Days, other than Study Day 1, on which no signs are observed are recorded on the Log of Animal Observations) Species

Sex

Route

Dose Level

Animals Coded*

Study Day OBSERVATIONS: Time Date

Date Dosed Page of NOTES: *An. Code

An. ID

No Signs Observed Reduced Motor Activity Ataxia Lost Righting Reflex

*Animal Code for Recording Observations

Convulsions ( ) Mydriasia

Read and Understood

Date

DEATH Observer

Figure 11.4

Example of a form for recording clinical observations in acute systemic toxicity studies.

Chapter 11:

Body Weight Considerations Body weight and feed consumption are frequently determined parameters in toxicity testing. To an extent, the ability of an animal to gain or maintain weight may be considered a sensitive, but nonspecific, indicator of health. While this is true in subchronic or chronic studies, its relevance in acute studies must be carefully considered. In most protocols, body weights are determined on Day 1 (prior to dosing), Day 7, and Day 14, which are the days mandated by most regulatory guidelines. Despite being common, the design is not well founded: if an animal has not died within seven days post-dosing, it has probably recovered and its body weight may not be noticeably different from controls by Day 14. A complete protocol addresses this problem by specifying more frequent body weight determinations (daily for the first three to five days of the observation period) so that not only can initial decreases (if they occur) be detected, but recovery can also be charted. Feed consumption measurements should be made at the same time, because it is difficult to determine the causes behind body weight changes in the absence of feed consumption data. Body weight loss accompanied by normal feed consumption implies something very different than body weight loss (or lack of gain) accompanied by lack of feed consumption. In the absence of feed consumption data, however, changes in body weight should still be considered indicative of a change in an animal’s health status. Yet another reason why body weight determinations are of questionable value in acute studies has to do with the statistical analysis of the data. Deaths may substantially alter group size and complicate analysis. The death of two of five animals causes a 40% decrease in group size and a substantial diminution of the power of any statistical test. In addition, the resulting data sets are censored: comparisons will often be between the control group, a dosage group where all the animals survive, and a high-dosage group where less than 50% of the animals survive to the end of the observation period. One has to question the utility of body weight changes if they occur at dosages that are acutely lethal. Body weight changes tended to occur only at dosages that were acutely lethal. Additionally, one would suspect that the censoring of body weights in groups where death occurs is not random; that is, the animals that die are most likely those that are most sensitive, while those that survive are the most resistant or robust. This problem can be addressed by building exclusionary criteria into a protocol. For example, one could statistically analyze body weight data in groups that only had less than 50% mortality.

Factors That Can Affect Acute Tests Many investigations into the sources of variability in acute toxicity testing have been conducted, and these have been reviewed by Elsberry (1986). The factors causing the greatest interstudy variation included lack of specifications for sex, strain, age, and weight range. When clearly defined, detailed protocols were used, interlaboratory variation was found to be minimal. Hence, it is equally important that the details of the protocol be well described and followed. It is not appropriate to draw dosage–response conclusion by comparing groups that differ substantially in age or that have been fed, fasted, or otherwise manipulated differently. Guidelines for standardization of acute toxicity testing were proposed by the interagency regulatory liaison group (IRLG, 1981; Elsberry, 1986). These do not differ markedly from those

Acute Systemic Toxicity Testing Device Safety Evaluation

Table 11.3

Neuromuscular Screen Observations

Observation

Nature of data generateda

Locomotor activity Righting reflex Grip strength (forelimb) Body temperature Salivation Startle response Respiration Urination Mouth breathing Convulsions Pineal response Piloerection Diarrhea Pupil size Pupil response Lacrimation Impaired gait Stereotypy Toe pinch Tail pinch Wire maneuver Hind-leg splay Positional passivity Tremors Extensor thrust Positive geotropism Limb rotation

S/N S N N Q Q S S Q S Q Q S S Q Q S Q S S S N S S S Q S

147

Correlates to which neutral componentb M/C C/M M C P S/C M/P/C P/M S C Reflex P/C GI tract/P/M P/C P/C S/P M/C C S (surface pain; spinal reflex) S (deep pain) C/M P/M S/C M/C C/M C M/C

a Data quantal (Q), scalar (S), or interval (N). Quantal data are characterized by being of an either/or variety, such as dead/alive or present/absent. Scalar data allow one to rank something as less than, equal to, or greater than other values, but one cannot exactly quantitate the difference between such rankings. Interval data is continuous data where one can assign (theoretically) an extremely accurate value to a characteristic that can be precisely related to other values in a quantitative fashion. b Peripheral (P), sensory (S), muscular (M), or central (C).

mandated by the Toxic Substance Control Act of 1986 (Gad and Chengelis, 1999).

Number, Size, and Sex of Dosage Groups The precision with which lethality and signs of toxicity are described will depend on the number of groups (hence, dosages) and the number of animals in each group. Between 1940 and 1980, the standard was to use from four to six dosages with ten animals per dosage. The current emphasis is on limiting the number of animals used for acute testing, particularly with recognition of the limited value of “precise” lethality data (Gad and Chengelis, 1999). Retrospective analyses by DePass (1989) and Olson et al. (1990) have demonstrated that decreasing group size to two to three animals generally has little impact on overall study results. Hence, the number and size of dosage groups will depend, to an extent, on the methods of statistical analysis. The classic statistical methods for analyzing lethality data (or, indeed, any quantal dosage-response data) were published between 1930 and 1960 and have been extensively reviewed by Armitage and Allen (1959) and Morrison et al. (1968). These methods are mentioned here with regard to the demand they make on protocol design—specifically, the number of dosage groups, the spacing of the dosages, and the number of animals per dosage group. The probit and moving average methods are the most commonly used today.

148

Safety Evaluation of Medical Devices

In general, all methods of calculation and computation are more precise if the designs are symmetrical (i.e., the dosages are evenly spaced and the group sized are equal). The probit method, first developed by Bliss (1935, 1957) and later refined by Finney (1971, 1985), is considered to be the most precise, but it requires at least two groups of partial responses (i.e., mortality greater than 0, but less than 100%). This may require dosing more than three groups until this criterion is met. It also deals ineffectively with groups that had either 0 or 100% mortality. (The most common correction for these groups is to substitute 0.1% for 0% and 99.7% for 100%.) The moving average method, first described by Thompson and Weil (1952), does not require partial responses, deals effectively with complete responses, and, therefore, can produce an acceptable estimate of an LD50 with as few as three groups of three to five animals each. The moving average method can also be used to design the experiment. Groups can be dosed in a sequential fashion as in a pyramiding study, with each step dictated by the moving average method. Once evidence of toxicity is observed, further dosing is discontinued. This method requires that the dosages be separated by a constant geometric factor (e.g., 2, 4, and 8 mg kg−1) and that groups be of equal size. Weil (1952), and later Gad (1999, 2001) has published tables that allow for the easy calculation of the LD50 using K = 3 (where K = the number of dosage groups minus one). The LD50 for K < 3 can be easily calculated without the aid of tables. In addition, methods for estimating the confidence limits of this calculated LD50 have also been published (Gad, 1999). Traditionally, the moving average method has not been extensively used because, while it yielded an estimate of the LD50, it did not given the slope of the (probit transformed) lethality curve. However, Weil (1983) has published a method for calculating a slope from the same data. Hence, an estimate of the LD50 and slope can be obtained from as few as three groups of three to five animals per group, provided that at least one group shows a response less than 50% and another shows a response greater than 50%. The Litchfield and Wilcoxon (1949) plotting method was once commonly used. It is certainly a valid method, and it poses no more restrictions on study design than those imposed by the probit method. The Litchfield–Wilcoxon method has become a victim of technology as modern, hand-held calculators and the ready availability of simple computer programs have made other methods more convenient to run. However, at least one software company has adopted the Litchfield– Wilcoxon method for its acute toxicity protocol package. The normit-chi square, developed by Berkson (1955), is also sometimes used. Like the probit method, the normit-chi square does not absolutely require equally spaced dosages or equal group sizes, but it does require at least one partial response. Hence, fewer dosage groups may be needed with the normit-chi square method than with the probit method. According to Waud (1972), the correction for including complete responses is better than that used for probit analysis but is still “tainted.” His method supposedly deals adequately with complete responses, but it is extremely complex and, probably for this reason, is rarely used. In an early paper, Karber (1931) published a simple method (often described but rarely cited) for calculating LD50. It does not require that dosages be equally spaced, but widely divergent dosages will lead to a biased result. The method was originally described for groups of equal size, but groups of slightly varying sizes can be used, provided they do not differ by more than a few animals each. In this

case, mean group size can be inserted into Karber’s formula with little change in accuracy. The formula is very simple, and one can calculate an acceptable estimate of the LD50 quickly with only a few arithmetic computations. This method, unlike those mentioned above, does not allow for calculating the confidence limit or slope of the probit response curve. Hence, if these calculated parameters are not sought, the Karber method allows one a bit more freedom in picking dosages. While much has been written about the influence of gender on acute lethality, most authors now agree that there are seldom any substantial differences in the LD50 due to sex (DePass et al., 1984; Gad and Chengelis, 1999). In those instances where there is a sex-related difference, females tend to be more sensitive than males (approximately 85% of the time). If one is willing to accept this amount of uncertainty, only one sex needs to be tested. Alternatively, as few as two to three animals per sex per dosage can be used. Schutz and Fuchs (1982) have demonstrated that, by pooling sexes, there are seldom any substantial differences in the LD50 calculations between groups consisting of five per sex versus three per sex. If there are no substantial differences between sexes (i.e., 70% mortality for males and 80% for females at a dosage), the results from pooling the sexes can provide a pooled LD50. For most safety considerations, LD50 derived on this basis will be acceptable, and will result in the use of fewer animals. The number of animals and the number of dosages are highly dependent on the type of data gathered. A few rules of thumb should be followed: (1) keep it lean: each additional group, animal, or test article added to a protocol makes the study exponentially more difficult to conduct— simplicity is one of the most important features of a screen; (2) the more parameters examined, the fewer the number of animals required; (3) if normal limits of a test parameter are relatively broad (e.g., serum glucose), more animals will be required than if the parameter is normally tightly controlled (e.g., prothrombin time). In general, three is the minimum and ten is the maximum number of animals required per group. Further, if a single chemical is examined per study, no more than three groups will be required. If more than one chemical is included in the study, then a single dosage (limit) group per chemical is the best design. Strictly speaking, an acute toxicity study is conducted to examine the effect of a single dose of a single compound. In designing specific toxicity screens, however, deviation from this principle is permissible if it increases screen sensitivity. For example, the sensitivity of mice to many indirect hepatotoxins will be enhanced by prior treatment with phenobarbital. Hence, the sensitivity of a hepatotoxicity screen will be enhanced if the mice are pretreated for three days with phenobarbital. The screen should be validated for consistency of response with both positive and negative control articles. A positive control article is one that is known to reliably produce the toxic syndrome the screen is designed to detect. Concurrent control groups are not required with each replicate. Rather, control groups should be evaluated on some regular basis to ensure that screen performance is stable. Because a screen relies on a biological system, it is not a bad idea to test the control benchmarks, particularly the positive ones, on a routine period basis. Not only does that give one increased confidence in the screen, but it also provides a historical base against which to compare the results of new test articles. Zbinden and colleagues refer to the positive control as the reference compound, and they have discussed some of

Chapter 11:

the general criteria to be applied in the selection of these compounds (Zbinden et al., 1984). Any changes to the design should trigger revalidation. Any analytical methods should be subjected to PASS (precision, accuracy, sensitivity, and selectivity) validation. Interpretation of specific toxicity screen data is not as straightforward as that of a general toxicity screen. This is because the data will often be continuous, following a Gaussian, or normal, distribution. This has two ramifications. First, for results around the threshold, it may be very difficult to differentiate between positive and negative responses. Second, for any one parameter, there is a real chance of false statistical significance (type I errors), especially if small numbers of animals are used. This occurrence is one of the reasons why specific toxicity screens should include the determination of more than one variable, since it is unlikely for multiple false positives to occur in the same group of animals. An undetected false positive could lead to the dropping of a promising candidate in error. False negatives, by contrast, may not be as critical (other than the time lost and the resources spent), because extensive subsequent tests should lead to the more complete description of the test article’s toxic properties. The problems described in the preceding paragraph assume that the screen will include a traditional (“negative,” or vehicle) control group, and that the data from the treated groups will be compared to those of the control group by standard methods. These problems will be minimized if no control group and, therefore, no traditional statistical comparisons are included. In addition, a decrease in the number of animals used simplifies the study. Data can be interpreted by comparison to a historical control data base, as described by Zbinden (1984). The threshold, or test criterion Xc, is calculated according to the following formula: Xc = m + (z)s where m is the population mean, s is the standard deviation, and z is an arbitrary constant. This formula is essentially a method of converting continuous data to quantal data: it is used to determine if individual animals are over the test threshold, not if the group mean is over the threshold. Analysis of screening data by comparison to experience (i.e., historical control data) and an activity criterion are discussed in greater detail in Chapter 4 and by Gad (2001). The higher the z value, the lower the probability of a false positive, but the lower the sensitivity of the screen. Again, including multiple parameters in the screen helps alleviate this problem. Zbinden has proposed a ranking procedure in which various levels of suspicion (LOS) or a level of certainty (LOC) is assigned to the result of a toxicity screen. This is simply a formalized fashion of stating that the more animals that respond, and the greater the severity of the response, the more certainty one has in drawing a conclusion. If relative comparisons are being made, this system provides a framework for ranking test articles and selecting those to continue to the next tier of testing. With regard to specific toxicity screening, behavioral toxicity screening is an area currently generating a great deal of interest. As reviewed by Hopper (1986), there are several reasons for this interest. First, the Toxic Substance Control Act of 1976 legislatively recognized behavioral measures as essential to examining chemicals for neurotoxic potential. Second, the structure and function of the CNS are not amenable to traditional methods of examination, in that profound behavioral changes can be induced in the absence of

Acute Systemic Toxicity Testing Device Safety Evaluation

149

any detectable morphological lesions. This large and somewhat controversial subject is outside the scope of this chapter. Specific screening strategies are presented and critically discussed by Hopper (1986). Other recommended references to consult for different perspectives on acute toxicity testing are Rhodes (2002), Brown (1980), and Arnold et al. (1990).

REFERENCES Armitage, P. and Allen, J. (1959). Methods of estimating the LD50 in quantal response data. J. Hygiene, 48: 298–322. Arnold, D.L., Grice, H.C. and Krewski, D.R. (1990). Handbook of in Vivo Toxicity Testing. Academic Press, San Diego, CA. Auletta, C. (1998). Acute systemic toxicity testing. In: S. Gad (Ed.), Handbook of Product Safety Assessment, 2nd Edn. Marcel Dekker, New York. Balazs, R. (1970). Measurement of acute toxicity. In: G. Paget, (Ed.), Methods of Toxicology. F.A. Davis Co., Philadelphia, pp. 49–81. Balazs, R. (1976). Assessment of the value of systemic toxicity studies in experimental animals. In: M., Mahlman, R., Shapiro, and H. Blumenthal, (Eds.), Advances in Modern Toxicity, Vol. 1, Part 1: New Concepts in Safety Evaluation. Hemisphere Publishing, Washington, D.C., pp. 141–153. Behrens, B. (1929). Evaluation of Digitalis leaves in frog experiments. Arch. Exp. Pathol. Pharmacol., 140: 236–256 (in German). Berkson, J. (1955). Estimate of the integrated normal curve by minimum normit chi-square with particular reference to bioassay. J. Am Stat. Assoc., 50: 529–549. Bliss, C. (1935). The calculation of the dosage mortality curve. Anal. Appl. Biol., 22: 34–167. Bliss, C. (1957). Some principles of bioassay. Am. Scientist, 45: 449–466. British Toxicology Society (1984). A new approach to the classification of substances and preparations on the basis of their acute toxicity. A report by the British Toxicology Society Working Party on Toxicology. Human Toxicol., 3: 85–92. Brown, V.K. (1980). Acute Toxicity. John Wiley & Sons, New York. Brown, V.K. (1983). Acute toxicity testing. In: M. Bals, R. Riddell and A. Worden (Eds.), Animals and Alternatives in Toxicity Testing. Academic Press, New York, pp. 1–13. Brownlee, K., Hodges, J. and Rosenblatt, M. (1953). The up-anddown method with small samples. J. Am. Stat., Assoc., 48: 262–277. Bruce, R. (1985). An up-and-down procedure for acute toxicity testing. Fund. Appl. Tox., 5: 151–157. Chan, P., O’Hara, G. and Hayes, A. (1982). Principles and methods for acute and subchronic toxicity. In: A. Hayes (Ed.), Principles and Methods of Toxicology. Raven Press, New York, pp. 1–51. Dayan, A. (1983). Complete program for acute toxicity testing—not only LD50 determination. Acta Pharmacol. Toxicol., 52 (Suppl 2): 31–51. Deichmann, W. and Gerarde, H. (1969). Toxicology of Drugs and Chemicals. Academic Press, New York. DePass, L. (1989). Alternative approaches in median lethality (LD50) and acute toxicity testing. Tox. Lett., 49: 159–170. DePass, L., Myers, R., Weaver, E. and Weil, C. (1984). An assessment of the importance of number of dosage levels, number of animals per dosage level, sex, and method of LD50 and slope calculations in acute toxicity studies. In: A. Goldberg (Ed.), Acute Toxicity Testing: Alternative Approaches, Vol. 2. Mary Ann Liebert, Inc., New York, pp. 139–153. Deichmann, W. and LeBlanc, T. (1943). Determination of the approximate lethal dose with about six animals. J. Ind. Hyg. Toxicol., 25: 415–417. Dixon, W. (1965). The up-and-down method for small samples. Amer. Statist. Assoc. J., 60: 967–978. Dixon, W.J. and Wood, A.M. (1948). A method for obtaining and analyzing sensitivity data. J. Am. Stat. Assoc., 43: 109–126. Elsberry, D. (1986). Screening approaches for acute and subacute toxicity studies. In: W. Lloyd (Ed.), Safety Evaluation of Drugs

150

Safety Evaluation of Medical Devices

and Chemicals. Hemisphere Publishing, Washington, D.C., pp. 145–151. Finney, D.J. (1971). Probit Analysis, Third Edn. Cambridge University Press, New York. Finney, D. (1985). The median lethal dose and its estimation. Arch. Toxicol., 56: 215–218. Fowler, J. and Rutty, D. (1983). Methodological aspects of acute toxicity testing particularly LD50 determinations: Present use in development of new drugs. Acta Pharmacol. Toxicol., 52 (Suppl. 2): 20–30. Fowler, J., Brown, J. and Bell, H. (1979). The rat toxicity screen. Pharmacol. Therapy, 5: 461–466. Fry, J., Gar, M. and Hammond, A. (1988). Choice of acute toxicity measures for comparison of in vivo/in vitro toxicity (short note). Alternatives to Laboratory Animals 16: 175–179. Gad, S. (1982). A neuromuscular screen for use in industrial toxicology. J. Toxicol. Envir. Health, 9: 691–704. Gad, S. and Chengelis, C.P. (1999). Acute Toxicology Testing, 2nd Edn. Academic Press, San Diego, CA. Gad, S. (2001). Statistics for toxicologists. In: A. Hayes (Ed.), Principles and Methods in Toxicology, 4th Edn. Taylor & Francis, Philadelphia. Gad, S., Smith, A., Cramp, A., Gavigan, F. and Derelanko, M. (1984). Innovative designs and practices for acute systemic toxicity studies. Drug Chem. Toxicol., 7: 423–434. Gad, S.C. (1999). Statistics and Experimental Design for Toxicologists, 3rd Edition, CRC Prem, Boca Raton, FL. Hazelette, J., Thompson, T., Mertz, B., Vuolo-Schuessler, L., Gree, J., Tripp, S., Robertson, P. and Triana, V. (1987). Rising dose tolerance (RDT) study: A novel scenario for obtaining preclinical toxicology/drug metabolism data. Toxicologist, 7, Abstract #846. van den Heuvel, M., Clark, D., Fielder, R., Koundakijian, P., Oliver, G., Pelling, D., Tomlinson. and Walker, A. (1990). The international validation of a fixed-dose procedure as an alternative to the classical LD50 test. Fd. Chem. Toxicol., 28: 469–482. Hopper, D. (1986). Behavioral measures in toxicology screening. In: W. Lloyd (Ed.), Safety Evaluation of Drugs and Chemicals. Hemisphere Publishing, New York, pp. 305–321. Interagency Regulatory Liaison Group (IRLG), Office of Consumer Affairs. (1981). Testing Standards and Guidelines Work Group (HFE88), Washington, D.C. Irwin, S. (1962). Drug screening and evaluative procedures. Science, 136: 123–128. Jenner, P., Hagan, E., Taylor, J., Cook, E. and Fitzhugh, O. (1964). Food flavorings and compounds of related structure. I. Acute oral toxicity. Fd. Cosmet. Toxicol., 2: 327–343. Karber, G. (1931). Contribution to the collective treatment of pharmacological serial experiments. Arch. Exp. Path. Pharmacol., 162: 480–483. LeBeau, J. (1983). The role of the LD50 determination in drug safety evaluation. Reg. Tox. Pharmacol., 3: 71–74. Litchfield, J. and Wilcoxon, F. (1949). A simplified method of evaluating dose-effect experiments. J. Pharmacol. Exp. Therap., 96: 99–113. Lörke, D. (1983). A new approach to practical acute toxicity testing. Arch. Toxicol., 54: 275–287. Malmfors, T., and Teiling, A. (1983). LD50—its value for the pharmaceutical industry in safety evaluation of drugs. Acta Pharmacol. Toxicol., 52 (suppl. 2): 229–246. McClain, R. (1983). Generating, interpreting and reporting information in toxicity studies. Drug Info. J., 17: 245–255. Morrison, J., Quinton, R. and Reinert, H. (1968). The purpose and value of LD50 determinations. In: E. Boyland and R. Goulding (Eds.), Modern Trends in Toxicology, Vol. I. Appleton-CenturyCrofts, London, pp. 1–17.

Muller, H. and Kley, H. (1982). Retrospective study on the reliability of an “approximate LD50” determined with a small number of animals. Arch.Toxicol., 51: 189–196. Office of Science Coordination, FDA (1984). Final Report on Acute Studies Workshop Sponsored by the U.S. Food and Drug Administration. U.S. Government Printshop, Washington, D.C. Olson, H., Fabian, R., Greener, Y., Pack, F., Zelinger, D. and Dean, J. (1990). Reduction in animals used for acute toxicity testing based on retrospective analysis. Toxicologist, 141 (abstract). Organization for Economic Co-operation and Development (OECD) (1991). OECD Guidelines for Testing of Chemicals, No. 401: Acute Oral Toxicity. OECD, Paris. Osterberg, R. (1983). Today’s requirements in food, drug, and chemical control. Acta Pharmacol. Toxicol., 52 (Suppl. 2): 201–228. Pendergast, W. (1984). Biological drug regulation. In: The Seventyfifth Anniversary Commemorative Volume of Food and Drug Law. Edited and Published by the Food and Drug Law Institute, Washington, D.C., pp. 293–305. Peterson, D., Ellis, Z., Harris, D. and Spadek, Z. (1987). Acute toxicity of the major alkaloids of cultivated Lupinus angustifolius seeds to rats. J. Appl. Toxicol., 7: 51–53. Piegorsh, W. (1989). Quantification of toxic response and the development of the median effective dose (ED50)—a historical perspective. Toxicol. Indust. Health. 5: 55–62. Rhodes, C. (2000). Principles of testing for acute toxic effects. In: B. Ballantyne, T. Marrs and T. Szversen (Eds.), General and Applied Toxicology. Macmillan References, Ltd., London, pp. 33–54. Rowan, A. (1981, April). The LD50 test: A critique and suggestions for alternatives. Pharmaceutical Technology, pp. 65–92. Schiantarelli, P. and Cadel, S. (1981). Piroxicam pharmacological activity and gastrointestinal damage by oral and rectal route: Comparison with oral indomethacin and phenylbutazone. Arzneim.-Forsch./Drug Res., 31: 87–92. Schutz, E. and Fuchs, H. (1982). A new approach to minimizing the number of animals in acute toxicity testing and optimizing the information of the test results. Arch. Toxicol., 51: 197–220. Smyth, H. and Carpenter, C. (1944). The place of the range-finding test in the industrial toxicology laboratory. J. Indus. Hyg. Tox., 26: 269–273. Sperling, F. (1976). Nonlethal parameters as indices of acute toxicity: Inadequacies of the acute LD50. In: M. Mehlman, R. Shapiro, and H. Blumenthal (Eds.), Advances in Modern Toxicology, Vol. 1, Part 1: New Concepts in Safety Evaluation. Hemisphere Publishing, Washington, D.C., pp. 177–191. Thompson, W. and Weil, C. (1952). On the construction of a table for moving average interpolation. Biometrics, 8: 51–54. Trevan, J. (1927). The error of determination of toxicity. Proc. Roy. Soc., B10: 483–514. Waud, D. (1972). Biological assays involving quantal responses. J. Pharmacol. Exp. Therap., 183: 577–607. Weil, C. (1952). Table for convenient calculation of median effective dose (LD50 or ED50) and instructions in their use. Biometrics 8: 249–263. Weil, C. (1983). Economical LD50 and slope determination. Drug and Chem. Tox., 6: 595–603. Zbinden, G. (1984). Acute toxicity testing, purpose. In: A. Goldberg (Ed.), Acute Toxicity Testing: Alternative Approaches. Mary Ann Liebert, New York, pp. 5–22. Zbinden, G. and Flury-Roversi, M. (1981). Significance of the LD50 test for the toxicological evaluation of chemical substances. Arch. Toxicol., 47: 77–99. Zbinden, G., Elsner, J. and Boelsterli, U. (1984). Toxicological screening. Regul. Toxicol. Pharmacol., 4: 275–286.

12 Genotoxicity

Genotoxicity encompasses all the potential means by which the genetic material of higher organisms may be damaged, with the potential for resulting serious consequences. Most forms of genotoxicity are expressions of mutagenicity—the induction of DNA damage and other genetic alterations, with binding to and changes in one or a few of DNA base pairs (gene mutations), Elastogenicity that is gross changes in chromosomal structure (i.e., chromosomal aberrations) or in chromosome numbers represents a different class of genotoxic changes, for which practical thresholds of effect are identifiable. It has been known for several hundred years that exposure to particular chemicals or complex mixtures can lead to cancer in later life (Doll, 1977), and it has been postulated more recently that chemicals can also induce heritable changes in man, leading to diseases in the next generation (ICEMC, 1983). There has been accumulating evidence that such changes can arise following damage to DNA and resulting mutations (see, e.g., Bridges, 1976). Therefore, it has become necessary to determine whether widely used chemicals or potentially useful new chemicals possess the ability to damage DNA. In industry, such information may be used to discard a new chemical drug or biomaterial if a safer alternative can be found, to control or eliminate human exposure for a genotoxic industrial compound or, for a drug, to proceed with development if benefits clearly outweigh risks. Data concerning the genotoxicity of a new material have become part of the basic biocompatibility information package. They are needed for decision-making and to reduce risks that might otherwise be unforeseen. ISO 10993-3 sets forth clear guidance on testing requirements, as summarized in Table 12.1.

Sixty-one codons code for specific amino acids (three produce stop signals), and as only 20 different amino acids are used to make proteins, one amino acid can be specified by more than one codon. The bases on one strand are connected together by a sugar (deoxyribose) phosphate backbone. DNA can exist in a single-stranded or double-stranded form. In the latter state, the two strands are held together by hydrogen bonds between the bases. Hydrogen bonds are weak electrostatic forces involving oxygen and nitrogen atoms. As a strict rule, one fundamental to mutagenesis, the adenine bases on one strand always hydrogen bond to the thymine bases on the sister strand. Similarly, guanine bases pair with cytosine bases. Adenine and thymine form two hydrogen bonds, and guanine and cytosine form three. Double-stranded DNA has a unique property in that it is able to make identical copies of itself when supplied with precursors, relevant enzymes, and cofactors. In simplified terms, two strands begin to unwind and separate as the hydrogen bonds are broken. This produces single-stranded regions. Complementary deoxyribonucleotide triphosphates then pair with the exposed bases under the control of a DNA polymerase enzyme. A structural gene is a linear sequence of codons which codes for a functional polypeptide, i.e., a linear sequence of amino acids. Individual polypeptides may have a structural, enzymatic or regulatory role in the cell. Although the primary structure of DNA is the same in prokaryotes and eukaryotes, there are differences between the genes of these two types of organism, in internal structure, numbers and mechanism of replication. In bacteria, there is a single chromosome, normally a closed circle, which is not complexed with protein, and replication does not require specialized cellular structures. In plant and animal cells, there are many chromosomes, each present as two copies, as mentioned earlier, and the DNA is complexed with protein. Replication and cell division require the proteinaceous spindle apparatus. The DNA of eukaryotic cells contains repeated sequences of some genes. Also, eukaryotic genes, unlike prokaryotic genes, have non-coding DNA regions called introns between coding regions called exons. This property means that eukaryotic cells have to use an additional processing step at transcription.

DNA STRUCTURE With the exception of certain viruses, the blueprint for all other organisms is contained in code by deoxyribonucleic acid (DNA), a giant macro-molecule whose structure allows a vast amount of information to be stored accurately. We have all arisen from a single cell, the fertilized ovum containing two sets of DNA (packaged with protein to form chromatin), one set from our mother, resident in the nucleus of the unfertilized ovum, the second set from our father via the successful sperm. Every cell in the adult has arisen from this one cell and (with the exception of the germ cell and specialized liver cells) contains one copy of these original chromosome sets. The genetic code is composed of four “letters”—two pyrimidine nitrogenous bases, thymine and cytosine, and two purine bases, guanine and adenine—which can be regarded functionally as arranged in codons (or triplets). Each codon consists of a combination of three letters; therefore, 43 (64) different codons are possible.

Transcription The relationship between the DNA in the nucleus and proteins in the cytoplasm is not direct. The information in the DNA molecule is transmitted to the protein-synthesizing machinery of the cell via another informational nucleic acid, called messenger RNA (mRNA), which is synthesized by an enzyme called RNA polymerase. Although similar to DNA, mRNAs are single-stranded, and possess the base uracil instead of thymine and the sugar ribose rather than deoxyribose. These molecules act as short-lived copies of the genes being expressed. 151

152

Safety Evaluation of Medical Devices

In eukaryotic cells, the initial mRNA copy contains homologues of both the intron and exon regions. The intron regions are then removed by enzymes located in the nucleus of the cell. Further enzymes splice the exon regions together to form the active mRNA molecules. In both groups of organisms mature mRNA molecules then pass out of the nucleus into the cytoplasm.

Excision Repair

Translation

(1) Preincision reactions. UvrA protein dimers are formed which bind to the DNA at a location distant from the damaged site. The UvrB protein then binds to the DNA–UvrA complex to produce an energy-requiring topological unwinding of the DNA via DNA gyrase. This area of unwinding is then translocated, again using ATP as an energy source, to the site of the damaged DNA. (2) Incision reactions. The UvrC protein binds to the DNA–UvrA, B complex and incises the DNA at two sites—seven bases to the 5¢ end and three bases to the 3¢ end of the damage. (3) Excision reactions. UvrD protein and DNA polymerase 1 excise the damaged bases and then resynthesize the strand, using the sister strand as a template. The Uvr complex then breaks down, leaving a restored, but nicked, strand. (4) Ligation reaction. The nick in the phosphate backbone is repaired by DNA ligase.

The next process is similar in both eukaryotes and prokaryotes, and involves the translation of mRNA molecules into polypeptides. This procedure involves many enzymes and two further types of RNA: transfer RNA (tRNA) and ribosomal RNA (rRNA). There is a specific tRNA for each of the amino acids. These molecules are involved in the transportation and coupling of amino acids into the resulting polypeptide. Each tRNA molecule has two binding sites, one for the specific amino acid, the other containing a triplet of bases (the “anticodon”) which is complementary to the appropriate codon on the mRNA. rRNA is complexed with protein to form a subcellular globular organelle called a ribosome. Ribosomes can be regarded as the “reading head” which allows the linear array of mRNA codons each to base-pair with an anticodon of an appropriate incoming tRNA/amino acid complex. The polypeptide chain forms as each tRNA/amino acid comes into register with them, RNA codon and with specific sites on the ribosome. A peptide bond is formed between each amino acid as it passes through the reading head of the ribosome (Venitt and Parry, 1984).

Some groups of enzymes (light-independent) are apparently organized to act cooperatively to recognize DNA lesions, remove them and correctly replace the damaged sections of DNA. The most comprehensively studied of these is the excision repair pathway. Briefly, the pathway can be described as follows:

A similar excision repair mechanism exists in mammalian cells (see, e.g., Cleaver, 1983). In both cases, the process is regarded as error-free and does not lead to the generation of mutations. However, this pathway can become saturated with excessive numbers of damaged DNA sites, forcing the cell to fall back on other repair mechanisms.

Gene Regulation Structural genes are regulated by a special set of codons, in particular “promoter” sequences. The promoter sequence is the initial binding site for RNA polymerase before transcription begins. Different promoter sequences have different affinities for RNA polymerase. Some sets of structural genes with linked functions have a single promoter and their coordinate expression is controlled by another regulatory gene called an operator. A group of such genes is called an operon. The activity of the operator is further controlled by a protein called a repressor, since it stops the expression of the whole operon by binding to the operator sequence, preventing RNA polymerase from binding to the promoter. Repressors can be removed by relevant chemical signals or in a time-related fashion. In the ways described above, only the genes required at a given moment are expressed. This not only helps to conserve the energy of the cell, but also is critical for correct cellular differentiation, tissue pattern formation, and formation of the body plan.

DNA Repair All living cells appear to possess several different major DNA repair processes (reviews: Walker, 1984; Rossman and Klein, 1988). Such processes are needed to protect cells from the lethal and mutating effects of heat-induced DNA hydrolysis; ultraviolet light; ionizing radiation; DNA reactive chemicals; free radicals, etc. In single-celled eukaryotes such as the yeast Saccharomyces cerevisiae, the number of genes known to be involved in DNA repair approaches 100 (Friedberg, 1988). The number in mammalian cells is expected to be at least equal to this and emphasizes the importance of correction of DNA damage.

Error-prone Repair Exposure of E. coli to agents or conditions that either damage DNA or interfere with DNA replication results in the increased expression of the so-called “SOS” regulatory network (review: Walker, 1984). Included in this network is a group of at least 17 unlinked DNA damage-inducible (din) genes. The din gene functions are repressed in undamaged cells by the product of the lexA gene (Little and Mount, 1982) and are induced when the LexA protein is cleaved by a process that requires modified RecA protein (RecA*), which then acts as a selective protease (Little, 1984). The din genes code for a variety of functions, including filamentation, cessation of respiration, etc. Included are the umuDC gene products, which are required for so-called “error-prone” or mutagenic DNA repair (Kato and Shinoura, 1977). The precise biochemical mechanism by which this repair is achieved is still not fully understood. Bacterial polymerase molecules have complex activities, including the ability to “proof-read” DNA—i.e., to ensure that the base-pairing rules of doublestranded DNA are met. It is hypothesized that Umu proteins may suppress this proof-reading activity, so that base mismatches are tolerated (Villani et al., 1978). Recent evidence suggests that DNA lesions are bypassed, and this bypass step required UmuDC proteins and RecA* protein (Bridges et al., 1987). The net result is that random base insertion occurs opposite the lesion which may result in mutation. Analogues of the umuDC genes can be found in locations other than the bacterial chromosome—e.g., plasmid pKM101 (Walker and Dobson, 1979), a derivative of the drug resistance plasmid R46 (Mortelmans and Stocker, 1979), which carried mucAB genes (Shanabruch and Walker, 1980)

Chapter 12:

(see pp. 879–880). Mutagenic repair, as controlled by umuDC, is not universal even among enterobacteria (Sedgwick and Goodwin, 1985). For instance, Salmonella typhimurium LT2 does not appear to express mutagenic repair (Walker, 1984). Thus, the usefulness of strains of this species is greatly enhanced by using derivatives containing plasmids with genes coding for error-prone repair (MacPhee, 1973; McCann et al., 1975).

Mismatch Repair Mispairs that break the normal base-pairing rules can arise spontaneously due to DNA biosynthetic errors, events associated with genetic recombination and the deamination of methylated cytosine (review: Modrich, 1987). With the latter, when cytosine deaminates to uracil, an endonuclease enzyme, N-uracil-DNA glycosylase (Lindahl, 1979), excises the uracil residue before it can pair with adenine at the next replication. However, 5-methyl cytosine deaminates to form thymine and will not be excised by a glycosylase. As a result, thymine exits on one strand paired with guanine on the sister strand, i.e., a mismatch. This will result in a spontaneous point mutation if left unrepaired. For this reason, methylated cytosines form spontaneous mutation “hot-spots” (review: Miller, 1985). The cell is able to repair mismatches by being able to distinguish between the DNA strand that exists before replication and a newly synthesized strand. The mechanism of strand-directed mismatch correction has been demonstrated in E. coli (see, e.g., Wagner and Meselson, 1976). In this organism, adenine methylation of d(G-A-T-C) sequences determines the strand on which repair occurs. Thus, parental DNA is fully methylated, while newly synthesized DNA is undermethylated, for a period sufficient for mismatch correction. By this means the organism preserves the presumed correct sequence—i.e., that present on the original DNA strand—and removes the aberrant base on the newly synthesized strand. Adenine methylation is achieved in E. coli by the dam methylase, which is dependent on S-adenosylmethionine. Mutants (dam) lacking this methylase are hypermutable, as would be expected by this model (Marinus and Morris, 1974).

The Adaptive Repair Pathway The mutagenic and carcinogenic effects of alkylating agents such as ethyl methane sulfonate are due to the generation of O6-alkylguanine residues in DNA, which result in point mutations. Bacterial and mammalian cells can repair a limited number of such lesions before DNA replication, thus preventing mutagenic and potentially lethal events taking place. If E. coli are exposed to low concentrations of a simple alkylating agent, a repair mechanism is induced that causes increased resistance to subsequent challenge with a high dose. This adaptation response was first described by Samson and Cairns (1977) and has recently been reviewed by Lindahl et al. (1988). The repair pathway is particularly well understood.

Plasmids Plasmids are extrachromosomal genetic elements that are composed of circular double-stranded DNA. In bacteria some can mediate their own transfer from cell to cell by conjugation—i.e., they contain a set of tra genes coding for tube-like structures, such as pili, through which a copy of plasmid DNA can pass during transfer.

Genotoxicity

153

Plasmids range in size from 1.5 to 200 million daltons. The number of copies per cell differs from plasmid to plasmid. Copy number relates to control of replication and this correlates with size—i.e., small plasmids tend to have large copy numbers per cell. This may relate to a lack of replication control genes (Mortelmans and Dousman, 1986).

Plasmids and DNA Repair Many plasmids are known to possess three properties: (1) increased resistance to the bactericidal effects of UV and chemical mutagens; (2) increased spontaneous mutagenesis; and (3) increased susceptibility to UV and chemically induced mutagenesis. Some plasmids possess all three properties; others may possess just one, e.g., increased susceptibility to mutagenesis (review: Mortelmans and Dousman, 1986). Often the profile of activity depends on the DNA repair status of the host cell (Pinney, 1980). Plasmid pKM101 carries DNA repair genes and has been widely used in strains used in bacterial mutagenicity tests.

Nature of Point Mutations The word “mutation” can be applied to point-mutations which are qualitative changes involving one or a few bases in base sequences within genes, as described below, as well as to larger changes involving whole chromosomes (and thus many thousands of genes), and even to changes in whole chromosome sets (described later under cytogenetics) (Figures 12.1–12.4). Point mutations can occur when one base is substituted for another (base substitution). Substitution of another purine for a purine base or of another pyrimidine for pyrimidine is called a transition, while substitutions of purine for pyrimidine or pyrimidine for purine are called transversions. Both types of base substitution have been identified within mutated genes. These changes lead to a codon change which can cause the “wrong” amino acid to be inserted into the relevant polypeptide and are known as mis-sense mutations. Such polypeptides may have dramatically altered properties if the new amino acid is close to the active center of an enzyme or affects the three-dimensional make-up of an enzyme or a structural protein. These changes, in turn, can lead to change or reduction in function, which can be detected as a change in phenotype of the affected cells. A base substitution can also result in the formation of a new inappropriate terminator (or non-sense) codon, and are thus known as non-sense mutations. The polypeptide formed from such mutated genes will be shorter than normal and is most likely to be inactive. Owing to the redundancy of the genetic code, about a quarter of all possible base substitutions will not result in an amino acid replacement and will be silent mutations. Bases can be deleted or added to a gene. As each gene is of a precisely defined length, these changes, if they involve a number of bases that is not a multiple of 3, result in a change in the “reading frame” of the DNA sequence and are thus known as frameshift mutations. Such mutations tend to have a dramatic effect on the polypeptide of the affected gene, as most amino acids will differ from the point of the insertion or deletion of bases onwards. Very often a new terminator codon is produced, so, again, short inactive polypeptides will result. Both types of mutation result in an altered polypeptide, which, in turn, can have a marked effect on the phenotype of the affected cell. Much use of phenotypic changes is made in mutagenicity tests.

154

Safety Evaluation of Medical Devices

ACTION: One nucleic acid is inserted (C is inserted) RESULT: All amino acids down the line are altered

C C mRNA amino acid translation

Figure 12.1

A G A

Glycine

T

Leucine

C

T

G C

Tyrosine

A C

Valine

A G A Serine

T

C C

Arginine

Mutation shift 1.

Base substitutions and frameshift changes occur spontaneously and can be induced by radiations and chemical mutagens. It is apparent that the molecular mechanisms resulting in these changes are different in each case, but the potential hazards associated with mutagens capable of inducing the different types of mutation are equivalent.

Suppressor Mutations In some instances a mutation within one gene can be corrected by a second mutational event at a separate site on the chromosome. As a result, the first defect is suppressed

and the second mutation is known as a suppressor mutation. Most suppressor mutations have been found to affect genes encoding for transfer RNAs. Usually the mutation causes a change in the sequence of the anticodon of the tRNA. Thus, if a new terminator or nonsense codon is formed as the first mutation, this can be suppressed by a second mutation forming a tRNA species that now has an anticodon complementary to a termination codon. Thus, the new tRNA species will supply an amino acid at the terminator site on the mRNA and allow translation to proceed. Surprisingly, most suppressors of this type do not adversely affect cell growth,

ACTION: One nucleic acid is inserted C is inserted) RESULT: All amino acids down the line are altered

C C A G A T C A T G C A C A G A T C C mRNA amino acid Glycine translation

Figure 12.2

Mutation shift 2.

Leucine

Valine

Arginine

Valine

STOP

Chapter 12:

Genotoxicity

155

ACTION: One nucleic acid is deleted (T is removed) RESULT: All amino acids down the line are altered

C C A G A T C A T G C A C A G A T C C mRNA Glycine amino acid translation

Figure 12.3

Leucine

Valine

Arginine

Valine

STOP

Mutation shift 3.

which implies that the cell can tolerate translation proceeding through termination signals, producing abnormal polypeptides. An alternative explanation is that the particular DNA sequences surrounding normal terminator codons result in a reduced efficiency of suppressor tRNAs (Bossi, 1985). Frameshift suppression is also possible. This can be achieved by a second mutation in a tRNA gene such that the anticodon of a tRNA molecule consists of 4 bases rather than 3, e.g., an extra C residue in the CCC anticodon sequence of a glycine tRNA gene. This change will allow correction of a +1 frameshift involving the GGG codon for glycine (Bossi, 1985).

Adduct Formation The earlier discussion of adaptive repair made reference to the fact that some unrepaired alkylated bases are lethal, owing to interference with DNA replication, while others, such as O6-methylguanine lead to mutation if unrepaired. These differences indicate that not all DNA adducts (i.e., DNA bases with additional chemical groups, not associated with normal DNA physiology) are equivalent. In fact, some adducts appear not to interfere with normal DNA functions or are rapidly repaired, others are mutagenic and yet others are lethal. Chemicals that form electrophilic species readily form DNA adducts. These pieces of information

ACTION: One nucleic acid is deleted (T is removed) RESULT: All amino acids down the line are altered

C C A G A C A mRNA amino acid Glycine translation

Figure 12.4

Mutation shift 4.

T G C A C A G A

Leucine Histidine

Valine

Serine

T

C C

Arginine

156

Safety Evaluation of Medical Devices

are hard-won, and the reader is recommended to read reviews of the pioneering work of Brooks and Lawley (review: Lawley, 1989) summarizing work identifying the importance of DNA adduct formation with polycyclic hydrocarbons and the importance of “minor” products of base alkylation such as O 6 -methylguanine, and, in addition, the work of the Millers in linking attack of nucleophilic sites in DNA by electrophiles to mutagenesis and carcinogenesis (Miller and Miller, 1971). If a DNA adduct involves the nitrogen or oxygen atoms involved in base-pairing, and the adducted DNA is not repaired, base substitution can result. Adducts can be small, such as the simple addition of methyl or ethyl groups, or they can be very bulky, owing to reaction with multiringed structures. The most vulnerable base is guanine, which can form adducts at several of its atoms (e.g., N7, C8, O6, and exocyclic N2 ) (Venitt and Parry, 1984). Adducts can form links between adjacent bases on the same strand (intrastrand cross-links) and can form interstrand cross-links between each strand of double-stranded DNA). The induction of frameshift mutation does not necessarily require covalent adduct formation. Some compounds that have a flat, planar structure, such as particular polycyclic hydrocarbons, can intercalate between the DNA strands of the DNA duplex. The intercalated molecules interfere with DNA repair enzymes or replication and cause additions and deletions of base-pairs. The precise mechanism is still unclear, although several mechanisms have been proposed. Hot-spots for frameshift mutation often involve sections of DNA where there is a run of the same base—e.g., the addition of a guanine to a run of six guanine residues. Such information led to a “slipped mispairing” model for frameshift mutation (Streisinger et al., 1966; Roth, 1974). In this scheme single strand breaks allow one strand to slip and loop out one or more base-pairs, the configuration being stabilized by complementary base-pairing at the end of the single-stranded region. Subsequent resynthesis results ultimately in additions or deletions of base-pairs (Miller, 1985).

Mutations Due to Insertion Sequences The subject of mutations due to insertion sequences is reviewed in Cullum (1985). Studies of spontaneous mutation in E. coli detected a special class of mutations that were strongly polar, reducing the expression of downstream genes (Jordan et al., 1967). These genes mapped as point mutations and reverted like typical point-mutations. However, unlike point-mutations, mutagens did not increase their reversion frequency. Further studies showed that these mutations were due to extra pieces of DNA that can be inserted into various places in the genome. They are not just random pieces of DNA but are “insertion sequences”, 0.7–1.5 kilobases long, that can “jump” into other DNA sequences. They are related to transposons, which are insertion sequences carrying easily detected markers such as antibiotic resistance genes, and Mu phages (bacterial viruses).

The Link Between Mutation and Cancer The change in cells undergoing normal, controlled cell division and differentiation to cells that are transformed, dividing without check, and are undifferentiated or abnormally differentiated, does not appear to occur as a single step—i.e., transformation is multistage. Evidence for this comes from in vitro studies, animal models and clinical observations—in particular, the long latent period between exposure to a carcinogen and the appearance of a tumor in

the target tissue. There is much evidence for the sequence of events shown in Table 12.1 (tumor initiation, promotion, malignant conversion and progression). Such a scheme provides a useful working model but clearly does not apply to all “carcinogens” in all circumstances. Study of Table 12.1 shows that there are several points where genetic change appears to play a role. Such change may occur spontaneously, due to rare errors at cell division such as misreplication of DNA or spindle malfunction, or may be induced by exposure to viruses (e.g., acute transforming retroviruses), ionizing and non-ionizing radiations absorbed by DNA (e.g., X-rays; UVC) or particular chemical species capable of covalently interacting with DNA (as discussed earlier) or with vital proteins, such as tubulin, that polymerize to form the cell division spindle apparatus.

Genotoxic Versus Non-genotoxic Mechanisms of Carcinogenesis The previous discussions of oncogene activation and human DNA repair deficiencies provide strong evidence for carcinogenesis via genotoxic mechanisms. However, it has been recognized for many years that cancers can arise without biologically significant direct or indirect interaction between a chemical and cellular DNA (see, e.g., Gatehouse et al., 1988). The distinction between non-genotoxic and genotoxic carcinogens has recently been brought into a sharper focus following the identification of a comparatively large number of “non-genotoxic” carcinogens by the United States National Toxicology Program (Tennant et al., 1987). These include a wide range of chemicals acting via a variety of mechanisms, including augmentation of high “spontaneous” tumor yields; disruption of normal hormonal homeostasis in hormone-responsive tissues; peroxisome proliferation; proliferation of urothelial cells following damage via induced kidney stones, etc. (Clayson, 1989). This author points out that a major effort is under way to determine whether many of these compounds can elicit similar effects in humans. Ashby and Tennant (1988) and Ashby et al. (1989) stress the significance of their observations that 16 tissues are apparently sensitive to genotoxic carcinogens, while a further 13 tissues are sensitive to both genotoxic and nongenotoxic carcinogens (Table 12.2). Also, genotoxic carcinogens tend to induce tumors in several tissues of both males and females in both rats and mice. This contrasts with nongenotoxic carcinogens, which may induce tumors at high doses, in one tissue, of one sex, of one species. Although it is most unlikely that all non-genotoxic carcinogens will prove to be irrelevant in terms of human risk, it appears from the analysis above that a proportion of carcinogens identified by the use of near-toxic levels in rodent bioassays are of dubious relevance to the induction of human cancer. For further discussion, see Butterworth and Slaga (1987).

Genetic Damage and Heritable Defects Concern about the effects of radiations and chemicals on the human gene pool, and the resulting heritable malformations and syndromes, has steadily risen during this century. The recognition that changes in morphology would result from changes in the hereditary material due to mutations (from the Latin word mutare, to change), was adopted by de Vries following observations on the evening primrose, Oenothera (de Vries, 1901). Muller went on to demonstrate that X-rays could induce mutations in the germ cells of the fruit fly Drosophila melanogaster (Muller, 1927).

ISO Genotoxicity Guidance

Gene mutations (changes in the sequences of DNA that code for critical proteins or functions) have been correlated to carcinogenicity and tumorigenicity.

Physical damage to chromosomes (large ordered stretches of DNA in the nuclei of cells) or clastogenicity can lead to DNA damage, in turn leading to abnormal and/or carcinogenic growth of cells.

Gene mutations

Chromosomal aberrations

Chromosomal aberration assay assesses the potential for physical damage to the chromosomes of mammalian cells by a biomaterial.

Ames assay (4 Salmonella typhimurium bacterial strains and Escherichia coli) is a reverse mutation assay. A bacterial mutation event causes the bacteria to become histidine (a vital amino acid) independent. Normal bacteria will not survive in the absence of histidine. Hypoxanthine guanine phosphoribosyl transferase (HGPRT) is a forward mutation assay. Mammalian cells that have been exposed to a mutagen will survive in the presence of a toxic substance (6-tioguanine).

Unscheduled DNA synthesis

Tests meeting requirements

ISO 10993-3 (1993) Biological Evaluation of Medical Devices—Part 3: Tests for Genotoxicity, Carcinogenicity, and Reproductive Toxicity states that at least three in vitro tests, two of which use mammalian cells, should be used to test for three levels of genotoxic effects: DNA effects, gene mutations, and chromosomal aberrations. “Suitable cell transformation systems may be used for carcinogenicity prescreening.”

Damage to DNA (deoxyribonucleic acid) by a chemical or material may result in genotoxic effects such as mutations, which in turn may lead to carcinogenicity. Damage to DNA causes the cell to manufacture new DNA to compensate for the loss or damage. This can be assessed by evaluating the formation of newly synthesized DNA.

Significance of test

DNA effects

Genotoxic effect to be assessed for conformance with ISO 1099-3

Table 12.1

Chapter 12: Genotoxicity

157

158

Safety Evaluation of Medical Devices

Table 12.2

Tissues Sensitive to Genotoxic and/or Non-genotoxic Carcinogens

Tissues sensitive primarily to genotoxins

Tissues sensitive to both genotoxinsand non-genotoxins

Stomach Zymbal gland Lung Subcutaneous tissue Circulatory system Clitoral gland Skin Intestine/colon Uterus Spleen Tunica vaginalis Bile duct Ovary Haderian gland Preputial gland [Multiple organ sites]

Nose Mammary gland Pituitary gland Integumentary system Kidney Urinary bladder Liver Thyroid gland Hematopoietic system Adrenal gland Pancreas Seminal vesicle Urinary tract Lymphatic system

Reproductive Effects If a potent genotoxin is able to cross the placental barrier, it is very likely to interfere with differentiation of the developing embryo and thus possess teratogenic potential. Indeed, many of the better studied teratogens are also mutagenic (Kalter, 1977). However, mutagens form only one class of teratogens and a large proportion of teratogens are not mutagenic. Alternative mechanisms of teratogenesis include cell death, mitotic delay, retarded differentiation, vascular insufficiency, inhibited cell migration, etc. (Beckman and Brent, 1986). It is known that more foetal wastage and many spontaneous abortions arise as a result of the presence of dominant lethal mutations in the developing embryo, many of which appear to be due to major chromosomal damage. In addition, impairment of male fertility is also a consequence of exposure to mutagens.

The human gene pool is known to carry many deleterious genes acquired from preceding generations which result in numerous genetic disease. It is clear that these arise as a result of DNA changes affecting particular chromosomes or genes. They can be grouped as follows: (1) Chromosome abnormalities, small changes in either number or structure. (2) Autosomal dominant gene mutations, in which a change in only one copy of the pair of genes is sufficient for the condition to be expressed. (3) Autosomal recessive gene mutations in which both copies of a gene must be mutated for the trait to become manifest. (4) Sex-linked conditions, which may also be recessive or dominant, where the mutant gene is on an X chromosome and will be expressed at high frequency in males (XY) and at a much lower frequency in females (XX), if the gene acts in a recessive manner.

EXPOSURE

INITIATION

CYTOGENETICS There are various types of cytogenetic change which can be detected in chromosomes. These are structural chromosome aberrations (CAs); numerical changes which could result in aneuploidy; and sister chromatid exchanges (SCEs). Chromosome aberration assays are used to detect the induction of chromosome breakage (clastogenesis) in somatic or germinal cells by direct observation of chromosome damage during metaphase analysis, or by indirect observation of micronuclei. Chromosome damage detected in these assays is mostly lethal to the cell during the cell cycle following the induction of the damage. Its presence, however, indicates a potential to induce more subtle chromosome damage which survives cell division to produce heritable cytogenetic changes. Cytogenetic damage is usually accompanied by other genotoxic damage such as gene mutation.

PROMOTION

Phenotypic changes Neoantigens Ectopic hormones Enzyme abnormalities Increased proliferation Altered differentiation Endothelial-stimulating factors

Body surface

RADIATION VIRUS

C

H

EM

IC

AL

CARCINOGEN

(5) Polygenic mutations, in which the condition results from the interaction of several genes and may include an environmental component.

Deactivation Excretion Genetic change Activation

Selective clonal expansion

PROGRESSION

Phenotypic changes Invasion Metastatis Neoantigens Ectopic hormones Enzyme abnormalities Altered cell surface Altered cytoskeleton Chromosome abnormalities

Genetic change

Metastasis

Cell heterogenicity

Nucleus Metastasis

Inhibition NORMAL CELL

INITIATED CELL

EXPOSURE

PRENEOPLASTIC LESION

MALIGNANT TUMOR

LATENCY PERIOD

−1

Figure 12.5

CONVERSION

INITIATION

PROMOTION

CLINICAL CANCER −12.775 days

CONVERSION

PROGRESSION

Schematic representation of events leading to neoplasia. Source: adapted from Harris et al. (1987).

Chapter 12:

Cytogenetic Damage and Its Consequences Structural and numerical chromosomal aberrations in somatic cells may be involved in the aetiology of neoplasia and in germ cells can lead to perinatal mortality, dominant lethality or congenital malformations in the offspring (Chandley, 1981), and some tumors (Anderson, 1990). Chromosome defects arise at the level of the individual chromosome or at the level of the chromosomal set, so affecting chromosomal number.

Individual Chromosome Damage Damage to individual chromosomes consists of breakage of chromatids, which must result from a discontinuity of both strands of the DNA in a chromatid. How mutagens produce chromosome breakage is not totally understood, but DNA lesions which are not in themselves discontinuities will produce breakage of a chromosome as a consequence of their interference with the normal process of DNA replication. In haploid microorganisms and prokaryotes chromosome breaks are usually lethal, but in diploid eukaryotes this is not so. According to Bender et al., (1974), in these organisms chromosome breaks may reconstitute in the same order, probably as a result of an enzyme repair process, resulting in no apparent cytogenetic damage; they may remain unjoined as fragments, which could result in cell death at the next or following mitoses—if, for example, unrejoined fragments are introduced into the zygote via treated germ cells, the embryo may die at a very early stage from a dominant lethal mutation; or they may rejoin in a different order from the original one, producing chromosomal rearrangements. There are various types of chromosomal rearrangements: Reciprocal translocations can result from the exchange of chromosomal segments between two chromosomes and, depending on the position of the centromeres in the rearranged chromosomes, different configurations will result. (1) Asymmetrical exchanges arise when one of the rearranged chromosomes carries both centromeres and is known as dicentric while the other carries none and is acentric. The cell or zygote carrying this anomaly usually dies, death being caused by segregation difficulties of the dicentric or the loss of the acentric fragment at cell division. Such a translocation contributes to dominant lethality. (2) Symmetrical exchanges occur when each rearranged chromosome carries just one centromere. This allows the zygote to develop normally, but when such heterozygoes form germ cells at meiosis, about half of their gametes will be genetically unbalanced, since they have deficiencies and duplications of chromosomal material. The unbalanced gametes which survive produce unbalanced zygotes, which results in death shortly before and after birth, or congenital malformations. Centric fusions (Robertsonian translocations) involve the joining together of two chromosomes, each of which has a centromere at or near one end, to produce a single metacentric or submetacentric chromosome. When Robertsonian translocations are produced in a germ cell and result from breakage and rejoining in the short arms of the two chromosomes, as a consequence of loss of the derived acentric fragments, a genetic deficiency can result. Some Robertsonian translocations are able to survive but others pose a risk. In heterozygotes the two arms of the translocation chromosome may pair with the two separate

Genotoxicity

159

homologous chromosomes at meiosis but segregate in a disorderly manner. Some of the resultant germ cells lack copies (nullisomy) or carry two copies (disomy) of one or other of the two chromosomes involved, which results in monosomic or trisomic embryos. Monosomics die early but trisomic embryos, which carry three copies of a chromosome, can survive to birth or beyond. If chromosome 21 is involved in the translocation, it can form a translocation trisomy and produce inherited Down syndrome (this differs from non-disjunctional Down syndrome trisomy). Deletions and deficiencies are produced when two breaks arise close together in the same chromosome. The two ends of the chromosome join when the fragment between the breaks becomes detached. At the next cell division the unattached piece of chromosome is likely to be lost. Large delections may contribute to dominant lethality. Small delections are difficult to distinguish from point mutations. Deletions may uncover pre-existing recessive genes. If one gene that is essential for survival is uncovered, it can act as a lethal in a homozygote and as a partial dominant in a heterozygote. Inversions occur when two breaks occur in the same chromosome. The portion between them is detached and becomes reinserted in the opposite way to its original position, i.e., the gene order is reversed. This need not cause a genetic problem, but imbalanced gametes could result in congenital malformation or fetal death.

Chromosome Set Damage Accuracy of chromosome replication and segregation of chromosomes to daughter cells requires accurate maintenance of the chromosome complement of a eukaryotic cell. Chromosome segregation in meiosis and mitosis is dependent upon the synthesis and functioning of the proteins of the spindle apparatus and upon the attachment and movement of chromosomes on the spindle. The kinetochores attach the chromosomes to the spindle and the centrioles are responsible for the polar orientation of the division apparatus. Sometimes such segregation events proceed incorrectly and homologous chromosomes separate, with deviations from the normal number (aneuploidy) into daughter cells or as a multiple of the complete karyotype (polyploidy). When both copies of a particular chromosome move into a daughter cell and the other cell receives none, the event is known as non-disjunction. Aneuploidy in live births and abortions arises from aneuploid gametes during germ cell meiosis. Trisomy or monosomy of large chromosomes leads to early embryonic death. Trisomy of the smaller chromosomes allows survival but is detrimental to the health of an affected person—e.g., Down syndrome (trisomy 21), Patau syndrome (trisomy 13), and Edward syndrome (trisomy 18). Sex chromosome trisomies (Klinefelter’s and XXX syndromes) and the sex chromosome monosomy (XO), known as the Turner syndrome, are also compatible with survival. Aneuploidy in somatic cells is involved in the formation of human tumors. Up to 10% of tumors are monosomic and trisomic for a specific chromosome as the single observable cytogenetic change. Most common among such tumors are trisomy 8, 9, 12, and 21 and monosomy for chromosomes 7, 22, and Y.

Test Systems In vivo and in vitro techniques are available to test mutagenic properties to demonstrate presence or lack of ability of the

160

Safety Evaluation of Medical Devices

Table 12.3

Table 12.4

Fifteen Common Assays Described by OECD

Assays for gene mutations Salmonella typhimurium reverse mutation assay (Ames test, bacteria) (OECD 471) Escherichia coli reverse mutation assay (bacteria) (OECD472) Gene mutation in mammalian cells in culture (OECD 476) Drosophila sex-linked recessive lethal assay (fruit fly) (OECD 477) Gene mutation in Saccaromyces cerevisiae (yeast) (OECD 480) Mouse spot test (OECD 484) Assays for Chromosomal and Genomic Mutations In vitro cytogenetic assay (OECD 473) In vivo cytogenetic assay (OICD 475) Micronucleus test (OECD 474) Dominant lethal assay (OECD 478) Heritable translocation assay (OECD 485) Mammalian germ cell cytogenetic assay (OECD 483) Assays for DNA Effects DNA damage and repair: unscheduled DNA synthesis in vitro (OECD 482) Mitotic recombination in Saccharomyces cerevisiae (yeast) (OECD 481) In vitro sister chromatid exchange assay (OECD 479)

In vitro

In vivo

✓ ✓ ✓

Genotoxicity Tests Recommended by ISO

Test

Mutation

Cell Type

Method

Salmonella reverse mutation assay (OECD 471) In vitro cytogenetic assay (OECD 473) Gene mutation in mammalian cells (OECD 476)

Gene

Bacterial

In vitro

Chromosome

Mammalian

In vitro

Gene

Mammalian

In vitro

✓ ✓ ✓

✓ ✓ ✓ ✓ ✓ ✓

• gene mutation in cultured mammalian cells such as Chinese Hamster V79 cell/HGPRT mutation system. ISO 10993 specifically requires three genotoxicity assays for all devices. The assays should preferably evaluate DNA effects, gene mutations and chromosomal aberrations; and two of the assays should preferably use mammalian cells. Guidance for providing tests for selection to meet these needs are the OECD guidelines, which include eight in vitro and seven in vivo assays.

ISO Test Profile ✓ ✓ ✓

test material to cause mutation or chromosomal damage or cause cancer, as summarized in Table 12.3. The material intended for intimate contact and long exposure should not have any genotoxic properties. The presence of unpolymerized materials and traces of monomers, oligomers, additives or biodegration products can cause mutations. Mutation can be a point mutation or chromosomal rearrangement caused by DNA damage. Therefore, the material’s ability to cause point mutation, chromosomal change, or evidence of DNA damage are tested. As we have seen, correlations exist between mutagenic and carcinogenic properties. Most carcinogens are mutagens, but not all mutagens are human carcinogens. The Ames Salmonella/Microsome Test is a principal sensitive mutagen screening test. Compounds are tested on the mutants of Salmonella typhimurium for reversion from a histidine requirement back to prototrophy. A positive result is seen by the growth of revertant bacteria. A microsomal activation system should be included in this assay. The use of all five bacterial test strains are generally required. Two non-bacterial mutagenicity tests are generally required to support the lack of mutagenic or carcinogenic potential. Some well known tests are:

ISO 10993 Part 3: Tests for Genotoxicity, Carcinogenicity and Reproductive Toxicity suggests that a series of three in vitro assays be conducted, at least two of which should use mammalian cells as a target (Table 12.4). The tests should address the three types of genotoxic effects: (1) gene mutations, (2) chromosomal and genomic aberrations, and (3) DNA effects. Three tests that are recommended are shown below (note that none of the three recommended tests assays for DNA effects!) In ISOs opinion, a profile of three in vitro genotoxicity tests is considered sufficient to establish safety for most medical devices; in vivo testing need only be done if in vitro tests are positive.

ICH Test Profile The International Conference on Harmonization recommends a rather different profile of genotoxicity tests for drugs (Table 12.5). There is a requirement to have an in vivo test conducted, but such is not needed prior to human exposure unless one or more of the in vitro studies yields a positive response.

In Vitro Test Systems The principal tests can be broadly categorized into microbial and mammalian cell assays. In both cases the tests are carried out in the presence and absence of in vitro metabolic activation enzymes, usually derived from rodent liver.

In Vitro Metabolic Activation The target cells for in vitro mutagenicity tests often possess a limited (often overlooked) capacity for endogenous metabolism of xenobiotics. However, to simulate the

Table 12.5

• the L5178Y mouse lymphoma test for mutants at the TK locus; • the induction of recessive lethals in Drosophilia melanogaster; • metaphase analysis of cultured mammalian cells and of treated animals; • Sister Chromatid Exchange Assay; • Unscheduled DNA Synthesis Assay; • Cell Transformation Assay; • SOS Repair System Assay; and

Genotoxicity tests recommended by ICH

Genotoxicity test—ICH

Mutation

Cell type

Method

A test for gene mutation in bacteria In vitro cytogenetic assay using mouse lymphomas tk cells In vivo test for chromosomal damage using rodent hematopoietic cells

Gene

Bacterial

In vitro

Chromosome

Mammalian

In vitro

Gene

Mammalian

In vivo

Chapter 12:

complexity of metabolic events that occur in the whole animal, there is a critical need to supplement this activity.

Choice of Species A bewildering variety of exogenous systems have been used for one purpose or another in mutagenicity tests. The choice begins with plant or animal preparations. The attraction of plant systems has stemmed from a desire to avoid the use of animals, where possible, in toxicity testing. In addition, plant systems have particular relevance when certain chemicals are being tested, e.g., herbicides. If animal systems are chosen, preparations derived from fish (see, e.g., Kada, 1981) and birds (Parry et al., 1985) have been used. However, by far the most widely used and validated are those derived from rodents—in particular, the rat. Hamsters may be preferred as a source of metabolizing enzymes when particular chemical classes are being screened—e.g., aromatic amines, heterocyclic amines, N-nitrosamines, and azo dyes (Prival and Mitchell, 1982; Haworth et al., 1983).

Choice of Tissue The next choice is that of source tissue. Preparations derived from liver are the most useful, as this tissue is a rich source of mixed-function oxygenases capable of converting procarcinogens to genetically active electrophiles. However, many extrahepatic tissues (e.g., kidney, lung, etc.) are also known to possess important metabolic capacity which may be relevant to the production of mutagenic metabolites in the whole animal.

Cell-free versus Cell-based Systems Most use has been made of cell-free systems—in particular, crude homogenates such as 9000 g supernatant (S9 fraction) from rat liver. This fraction is composed of free endoplasmic reticulum, microsomes (membrane-bound packets of “membrane-associated” enzymes), soluble enzymes and some cofactors. Hepatic S9 fractions do not necessarily completely reflect the metabolism of the whole organ, in that they mainly possess phase I metabolism (e.g., oxygenases) and are deficient in phase II systems (e.g., conjugation enzymes). The latter are often capable of efficient detoxification, while the former are regarded as “activating.” This can be a strength in that S9 fractions are used in screening tests as a surrogate for all tissues in an animal, some of which may be exposed to reactive metabolites in the absence of efficient detoxification. Many carcinogens are organ-specific in extrahepatic tissues, yet liver S9 fraction will reveal their mutagenicity. The deficiency of S9 fractions for detoxification can also be a weakness, in that detoxification may predominate in the whole animal, such that the potential carcinogenicity revealed in vitro is not realized in vivo. Cell-free systems, when supplemented with relevant cofactors, are remarkably proficient, despite their crudity in generating reactive electrophiles from most procarcinogens. However, they provide at best a broad approximation of in vivo metabolism and can fail to produce sufficient quantity of a particular reactive metabolite to be detectable by the indicator cells or they can produce inappropriate metabolites that do not play a role in vivo (Gatehouse and Tweats, 1987, for discussion). Some of these problems can be overcome by the use of cell-based systems—in particular, primary hepatocytes. Hepatocytes closely simulate the metabolic systems found in the intact liver and do not require additional cofactors for

Genotoxicity

161

optimal enzyme activity. However, apart from greater technical difficulties in obtaining hepatocytes as opposed to S9 fraction, hepatocytes can effectively detoxify particular carcinogens and prevent their detection as mutagens. Despite these difficulties, hepatocytes have a role to play in mutagenicity screening, in both bacterial and mammalianbased systems (Tweats and Gatehouse, 1988).

Inducing Agents The final choice considered here is whether to use “uninduced” liver preparations or those derived from animals pretreated with an enzyme inducer to promote high levels of metabolic activity. If induced preparations are preferred, which inducer should be used? It appears that uninduced preparations are of limited use in screening assays, as they are deficient in particular important activities such as cytochrome P-450IA1 cytochrome oxygenases. In addition, species and organ differences are most divergent with uninduced enzyme preparations (Brusick, 1987). The above differences disappear when induced microsomal preparations are used. A number of enzyme inducers have been used, the most popular being Aroclor 1254, which is a mixture of polychlorinated bipheynyls (as described by Ames et al., 1975). However, concern about the toxicity, carcinogenicity and persistence of these compounds in the environment has led to the use of alternatives, such as a combination of phenobarbitone (phenobarbital) and β-naphthoflavone (5,6-benzoflavone). This combination results in the induction of a range of mono-oxygenases similar to that induced by Aroclor 1254 (see, e.g., Ong et al., 1980). More selective inducers such as phenobarbitone (cytochrome P-450IIa1, P-450IIB1) or 3-methylcholanthrene (cytochrome P-450IA1) have also been used. In summary, genetic toxicity tests with both bacterial and mammalian cells are normally carried out with rat liver cell-free systems (S9 fraction) from animals pretreated with enzyme inducers. However, investigations should not slavishly follow this regimen: there may be sound scientifically based reasons for using preparations from different species or different organs, or for using whole cells such as hepatocytes.

Standard Method of S9 Fraction Preparation The following method describes the production of hepatic S9 mix from rats induced with a combination of phenobarbitone and β-naphthoflavone, and is an adaptation of the method described by Gatehouse and Delow (1979). Male albino rats within the weight range 150–250 g are treated with phenobarbitone sodium 16 mg mL−1, 2.5 mL kg−1 in sterile saline, and β-naphthoflavone 20 mg mL−1 in corn oil. A fine suspension of the latter is achieved by sonicating for 1 hr. These solutions are dosed by intraperitoneal injection on days 1, 2, and 3. Phenobarbitone sodium is normally administered between 0.5 and 2 hr prior to β-naphthoflavone. The animals are killed on day 4 by cervical dislocation and the livers removed as quickly as possible and placed on ice-cold KCI buffer (0.01 M Na2HPO4 + KCI 1.15%). The liver is cleaned, weighted, minced and homogenized (in an Ultra Turrx homogenizer) in the above buffer to give a 25% (w/v) liver homogenate. The homogenate is stored at 4ºC until it can be centrifuged at 9000 g for 15 min. The supernatant is decanted, mixed and divided into 2 mL volumes in cryotubes. These are then snap-frozen in liquid nitrogen.

162

Safety Evaluation of Medical Devices

Storage at −196ºC for up to 3 months results in no appreciable loss of most P-450 isoenzymes (Ashwood-Smith, 1980). Quality control of S9 batches is usually monitored by the ability to activate compounds known to require metabolism to generate mutagenic metabolites. This is a rather crude approach and more accurate data can be obtained by measuring biochemical paramets—e.g., protein, cytochrome P-450 total activity (from crude S9), and related enzyme activities from purified microsomes) such as 7-ethoxyresorufin-0-deethylase and 7-methoxycoumarin-0-demethylase to give an indication of S9 batch-to-batch variation and to set standards for rejecting suboptimal batches (Hubbard et al., 1985). For further details on critical features affecting the use and limitations of S9 fraction, see Gatehouse and Tweats (1987).

S9 Mix The S9 fraction prepared as described above is used as a component in “S9 mix” along with buffers and various enzyme cofactors. The amount of S9 fraction in the S9 mix can be varied, but a “standard” level of 0.1 mL mL−1 of S9 mix (or 10% S9) is often recommended for general screening. No single concentration of S9 fraction in the S9 mix will detect all classes of genotoxic carcinogen with equal efficiency (Gatehouse et al., 1990). Some mutagens, including many polycyclic aromatic hydrocarbons, are activated to mutagens by higher than normal levels of S9 fraction in the S9 mix (see, e.g., Carver et al., 1985). The mixed-function oxidases in the S9 fraction require NADPH, normally generated from the action of glucose6-phosphate dehydrogenase acting on glucose-6-phosphate and reducing NADP, both of which are normally supplied as cofactors. As an alternative, isocitrate can be substituted for glucose-6-phosphate (to be used as a substrate by isocitrate dehydrogenase) (Linblad and Jackim, 1982). Additional cofactors may be added (e.g., flavin mononucleotide), when particular classes of compound such as azo dyes are being tested (Prival et al., 1984), or acetyl coenzyme A when aromatic animes such as benzidine are being tested (Kennelly et al., 1984). The composition of a “standard” S9 mix is given in Table 12.6.

Bacterial Mutation Tests The study of mutation in bacteria (and bacterial viruses) has had a fundamental role in the science of genetics in the twentieth century. In particular, the unraveling of

Table 12.6

Composition of Standard S9 Mix

Constituent Glucose-6-phosphate Nicotinamide adenine dinucleotide phosphate MgCl26H2O Salt KCl Phosphate buffer 90.2 M) Distilled water to make up to the required volume S9 fraction added at 0.1 mL per mL of S9 mix

Final conc. in mix (mM) 5 4 8 33 100

For assays using cultured mammalian cells, phosphate buffer and distilled water are replaced by tissue culture medium, as high concentrations of Na and K salts are toxic to such cells. The concentration of S9 fraction in the S9 mix varies, depending on the relevant assay (see individual sections). Once prepared, S9 mix should be used as soon as possible, and should be stored on ice until required. S9 fraction, once thawed, should not be refrozen for future use.

biochemical anabolic and catabolic pathways, the identification of DNA as the hereditary material, the fine structure of the gene, the nature of gene regulation, etc., have all been aided by bacterial mutants. As an offshoot of studies of genes concerned with the biosynthesis of amino acids, a range of E. coli (see, e.g., Yanofsky, 1971) and Salmonella typhimurium strains (see, e.g., Ames, 1971) with relatively well-defined mutations in known genes became available. Thus, bacteria already mutant at an easily detectable locus are treated with a range of doses of the test material to determine whether the compound can induce a second mutation that directly reverses or suppresses the original mutations. Thus, for amino acid auxotrophs, the original mutation has resulted in loss of ability to grow in the absence of the required amino acid. The second mutation restores prototrophy—i.e., the affected cell is now able to grow in the absence of the relevant amino acid, if provided with inorganic salts and a carbon source. This simple concept, in fact, underlines the great strength of these assays, for it provides enormous selective power which can identify a small number of the chosen mutants from a population of millions of unmutated cells and cells mutated in other genes. The genetic target—i.e., the mutated DNA bases in the gene in question (or bases in the relevant tRNA genes; see the discussion of suppressor mutations)—can thus be very small, just one or a few bases in length. An alternative approach is to use bacteria to detect “forward mutations.” Genetic systems which detect forward mutations have an apparent advantage, in that a wide variety of genetic changes may lead to a forward mutation—e.g., point mutation, deletions, insertions, etc. In addition, forward mutations in a number of different genes may lead to the same change in phenotype; thus, the genetic target is much larger than that seen in most reverse mutation assays. However, if a particular mutagen causes rare specific changes, these changes may be lost against the background of more common events (Gatehouse et al., 1990). Spontaneous mutation rates tend to be relatively high in forward mutation systems. Acquisition of resistance to a toxic chemical (e.g., an amino acid analogue or antibiotic) is a frequently used genetic marker in these systems. For instance, the use of resistance to the antibiotic streptomycin preceded the reversion assays in common use today.

Reversion Tests—Background There are several excellent references describing the background and use of bacteria for reversion tests (Brusick, 1987b; Gatehouse et al., 1990). Three different protocols have been widely used: plate incorporation assays, treat and plate tests, and fluctuation tests. These methods are described in detail in the following sections. Fundamental to the operation of these tests is the genetic compositions of the tester strains selected for use.

Genetic Make-up of Tester Strains The most widely used strains are those developed by Bruce Ames and colleagues which are mutant derivatives of the organism Salmonella typhimurium. Each strain carries one of a number of mutations in the operon coding for histidine biosynthesis. In each case the mutation can be reverted either by base-change or by frameshift mutations. The genotype of the commonly used strains is shown in Table 12.7.

Chapter 12:

Table 12.7 Genotype of Commonly Used Strains of Salmonella typhimurium LT2 and their Reversion Events. Strain

Genotypea

Reversion events

TA1535

hisG46 rfa ƒ gal chlD bio uvrB hisG46 ƒ rfa gal chlD bio uvrB (pKM101) hisC3076 ƒ rfa gal chlD bio uvrB hisD3052 ƒ rfa gal chlD bio uvrB hisD3052 ƒ rfa gal chlD bio uvrB pKM101) hisD6610 hisO1242 rfa ƒ gal chlD bio uvrB (pKM101) hisƒ (G)8476 rfa galE (pAQ1) (pKM101)

Subset of base-pair substitution events Subset of base-pair substitution events Frameshifts

TA100 TA1537 TA1538 TA98 TA97 TA102

Frameshifts Frameshifts Frameshifts All possible transitions and transversions; small deletions

The Use of the Plasmid pKM101 Salmonella typhimurium LT2 strains do not appear to possess classical “error-prone” repair as found in E. coli strains and some other members of the Enterobacteria (Walker, 1984; Sedwick and Goodwin, 1985). This is due to a deficiency in umu D activity in these Salmonella strains (Herrera et al., 1988; Thomas and Sedgewick, 1989). One way to overcome this deficiency and to increase sensitivity to mutagens is to use strains containing a plasmid carrying analogues to the umu DC genes.

E. coli Tester Strains Ames and colleagues have made an impressive contribution to mutagenicity testing by the development of the Salmonella/ microsome test and, in particular, its application in the study of environmental mutagens. In genetic terms, Salmonella strains are, in some ways, not the best choice (see, e.g., Venitt and Croften-Sleigh, 1981). Unlike the Salmonella strains, E. coli B strains such as the WP2 series developed by Bridges, Green and colleagues (Bridges, 1972; Green and Muriel, 1976) possess the umuDC+ genes involved in generating mutations; they are also part-rough and thus allow many large molecules to enter the cell. In addition to being effective general strains for mutagen detection, studies by Wilcox et al. (1990) have shown that a combination of E. coli WP2 trp E (pKM101), which has a functioning excision repair system for the detection of cross-linking agents, and E. coli WP2 trp E uvrA (pKM101) can be used as alternatives to Salmonella TA102 for the detection of oxidative mutagens. The E. coli strains have the advantage of lower spontaneous mutation rate and are somewhat less difficult to use and maintain.

Storage and Checking of Tester Strains Detailed instructions for maintenance and confirmation of the phonotypes of the various tester strains are given in Maron and Ames (1983) and Gatehouse et al. (1990). Permanent master cultures of tester strains should be stored in liquid nitrogen or in dry ice. Such cultures are prepared from fresh nutrient broth cultures, to which DMSO is added as a cryopreservative. These cultures are checked for the various characteristics before storage as described below. Cultures for use in individual experiments should be set up by inoculation from the master culture or from a plate made

Genotoxicity

163

directly from the master culture, not by passage from a previously used culture. Passage in this way will inevitably increase the number of pre-existing mutants, leading to unacceptably high spontaneous mutation rates (Gatehouse et al., 1990). The following characteristics of the tester strains should be confirmed at monthly intervals or if the internal controls of a particular experiment fail to meet the required limits: • Amino acid requirement. • Sensitivity to the lethal effects of the high-molecularweight dye crystal violet for those strains carrying the rfaE mutation. • Increased sensitivity to UV irradiation for those strains carrying the uvrA or uvrB mutations. • Resistance to ampicillin for strains carrying pKM101 and resistance to tetracycline for strains carrying pAQ1. • Sensitivity to diagnostic mutagens. This can be measured very satisfactorily by testing pairs of strains—one giving a strongly positive response, the partner a weak response. The importance of these checks, together with careful experiment-to-experiment controls of spontaneous mutation rates and response to reference mutation rates and response to reference mutagens, cannot be overstressed; failure to apply them can result in much wasted effort.

PLATE INCORPORATION ASSAY Protocol for Dose Ranging and Selection Before carrying out the main tests, it is necessary to carry out a preliminary toxicity dose ranging test. This should be carried out following the same basic protocol as the mutation test, except that instead of scoring the number of mutants on, for example minimal media plates with limiting amounts of a required amino acid, the number of survivors is scored on fully supplemented minimal media. A typical protocol is outlined below: (1) Prepare a stock solution of the test compound at a concentration of 50 mg mL−1 in an appropriate solvent. It may be necessary to prepare a lower concentration of stock solution, depending on the solubility of the test compound. (2) Make dilutions of the stock solution. (3) To 2.0 mL aliquots of soft agar overlay medium (0.6% agar and 0.5% sodium chloride in distilled water) containing a trace of histidine and excess biotin and maintained at 45ºC in a dry block, add 100 µL at a solution of the test article. Use only one plate per dilution. (4) Mix and pour on to dried Vogel and Bonner minimal medium plates as in an Ames test, including an untreated control and a solvent control, if necessary. The final concentrations of test compound will be 5000, 1500, 500, 150 and 50 µg plate −1. (5) Repeat step (3), using 0.5 mL of 8% S9 mix per 2.0 mL aliquot of soft agar in addition to the test compound and tester strain. The S9 mix is kept on ice during the experiment. (6) Incubate the plates for 2 days at 37ºC and examine the background lawn of growth with a microscope (58 eyepiece lens, 510 objective lens). The lowest concentration giving a depleted background lawn is regarded as a toxic dose. This test will also demonstrate excess growth, which may indicate the presence of histidine or tryptophan on their

164

Safety Evaluation of Medical Devices

precursors in the test material, which could make testing for mutagenicity impracticable by this method. When setting the maximum test concentration, it is important to test into the mg plate−1 range where possible (Gatehouse et al., 1990), as some mutagens are only detectable when tested at high concentrations. However, for non-toxic, soluble mutagens an upper limit of 5 mg plate−1 is recommended (DeSerres and Shelby, 1979). For less soluble compounds at least one dose exhibiting precipitation should be included.

Ames Salmonella/Plate Incorporation Method The following procedure is based on that described by Ames and colleagues (Maron and Ames, 1983), with additional modifications. (1) Each selected test strain is grown for 10 h at 37ºC in nutrient broth (Oxoid No. 2) or supplemented minimal media (Vogel-Bonner) on an orbital shaker. A timing device can be used to ensure that cultures are ready at the beginning of the working day. (2) 2.0 mL aliquots of soft agar overlay medium are melted just prior to use and cooled to 50ºC, and relevant supplements added—i.e., L-histidine, final concentration 9.55 µg mL−1, and D-biotin, 12 µg mL−1. (Note: If E. coli WP2 tester strains are used, the only supplement required is tryptophan 3.6 µg mL−1.) The medium is kept semi-molten by holding the tubes containing the medium in a hot aluminum dry block, held at 45ºC. It is best to avoid water baths as microbial contamination can cause problems. (3) The following additions are made to each tube of top agar: the test article (or solvent control) in solution (10–200 µL), the test strain (100 µL) and, where necessary, S9 mix (500 µL). The test is carried out in the presence and absence of S9 mix. The exact volume of test article or solvent may depend on toxicity or solubility, as described in the preceding section. (4) There should be at least three replicate plates per treatment with at least five test doses plus untreated controls. Duplicate plates are sufficient for the positive and sterility control treatments. The use of twice as many negative control plates as used in each treatment group will lead to more powerful tests from a statistical standpoint (Mahon et al., 1989). (5) Each tube of top agar is mixed and quickly poured onto dried prelabeled Vogel–Bonner basal agar plates. (6) The soft agar is allowed to set at room temperature and the plates are inverted and incubated (within 1 hr of pouring) at 37ºC in the dark. Incubation is continued for 2–3 days. (7) Before scoring the plates for revertant colonies, the presence of a light background lawn of growth (due to limited growth of non-revertant colonies before the trace of histidine or tryptophan is exhausted) should be confirmed for each concentration of test article by examination of the plate under low power of a light microscope. At concentrations that are toxic to the test strains, such a lawn will be deplated and colonies may appear that are not true revertants but surviving, non-prototrophic cells. If necessary, the phenotype of any questionable colonies (pseudo-revertants) should be checked by plating on histidine or tryptophan-free medium. (8) Revertant colonies can be counted by hand or with an automatic colony counter. Such machines are relatively

accurate in the range of colonies normally observed (although calibration against manual counts is a wise precaution). Where accurate quantitative counts of plates with large numbers of colonies and required, only manual counts will give accurate results.

Controls Positive Controls Where possible, positive controls should be chosen that are structurally related to the test article. This increases the confidence in the results. In the absence of structurally related mutagens, the set of positive controls given in Table 10.8 can be used. The use of such controls validates each test run and helps to confirm the nature of each strain. Pagano and Zeger (1985) have shown that it is possible to store stock solutions of most routinely used positive controls (sodium azide, 2-aminoanthracene, benzo[a]phyene, 4-nitroquinoline oxide) at −20ºC to −80ºC for several months, without loss of activity. This measure can help reduce potential exposure to laboratory personnel.

Untreated/Vehicle Controls Untreated controls omit the test article, but are made up to volume with buffer. The vehicle control is made up to volume with the solvent used to dissolve the test substance. It is preferable to ensure that each of the treated plates contains the same volume of vehicle throughout. As detailed by Gatehouse and Tweats (1987), the nature and concentration of solvent may have a marked effect on the test result. Dimethylsulfoxide is often used as the solvent of choice for hydrophobic compounds. However, there may be unforeseen effects, such as an increase in mutagenicity of come compounds—e.g., p-phenylenediamne (Burnett et al., 1982)—or a decrease in mutagenicity of others, such as simple aliphatic nitrosqamines (Yahagi et al., 1977). It is essential to use fresh batches of the highest purity grade available and to prevent decomposition/oxidation on storage. The products after oxidation, etc., are both toxic and can induce base-pair substitutions in both bacterial and mammalian assays. Finally, DMSO and other organic solvents can inhibit the oxidation of different premutagens by microsomal mono-oxygenases (Wolff, 1977a, b). To reduce the risk of artifactual results, it is essential to use the minimum amount of organic solvent (e.g., <2% w/w) compatible with adequate testing of the test chemical. It is important to keep a careful check of the number of mutant colonies present on untreated or vehicle control plates. These numbers depend on the following factors: (1) The repair status of the cell—i.e., excision repair-deficient strains tend to have more “spontaneous mutants” than repair-proficient cells. (2) The presence of mutator plasmids. Excision-deficient strains containing pKM101 have a higher spontaneous mutation rate at both base substitution and frameshift loci than excision-proficient strains. (3) The total number of cell divisions that take place of the cells in the supplemented top agar. This is controlled by the supply of nutrients—in particular, histidine. Rat liver extracts may also supply trace amounts of limiting nutrients, resulting in a slight increase in the spontaneous yield of mutants in the presence of S9 mix. (4) The size of the initial inoculum. During growth of the starting culture, mutants will arise. Thus, if a larger

Chapter 12:

starting inoculum is used, more of these “pre-existing” mutants will be present per plate. In fact, the “plate mutants” arising as described in point (3) predominate. (5) The intrinsic mutability of the mutation in question. In practice the control mutation values tend to fall within in relatively precise range for each strain. Each laboratory should determine the normal range of revertant colonies per plate for each strain. Deviations in background reversion counts from the normal range should be investigated. It is possible that cross-contamination, variations in media quality, etc., have occurred that may invalidate particular experiments. Frequent checks should also be made on the sterility of S9 preparations, media and test articles. These simple precautions can prevent loss of valuable time and resources.

Genotoxicity

165

variance is not approximately constant, then a transformation may be applied or a weighted analysis may be carried out. If the dose scale tends to a plateau, then the dose scale may be transformed. If counts decline markedly at high doses, then linear regression is inappropriate. (4) Dunnett’s method, perhaps with a transformation, is recommended when counts decline markedly at one or two high doses. However, when the dose response shows no such decline, other methods may be more powerful. (5) Wahrendorf’s non-parametric method avoids the complications of transformations of weighting and is about as powerful as any other method. However, it is inappropriate when the response declines markedly at high dose.

Preincubation Tests Evaluation of Results At least two independent assays are carried out for each test article. The criterion for positive response is a reproducible and statistically significant result at any concentration for any strain. When positive results are obtained, the test is repeated, using the strain(s) and concentration range with which the initial positive results were observed. This range may be quite narrow for toxic agents. Several statistical approaches have been applied to the results of plate incorporation assays (review: Mahon et al., 1989). These authors make a number of important suggestions to maximize the power of statistical analyses; those that relate to the method of analysis are reproduced below. (1) Unless it is obvious that the test agent has had no effect, the data should be plotted, to give a visual impression of the form of any dose response and the pattern of variability. (2) Three methods of analysis—linear regression (e.g., Steel and Torrie, 1960); a multiple comparison analysis, Dunnett’s method (Dunnett, 1955); and a non-parametric analysis, Wahrendorf’s method (Wahrendorf et al., 1985)—can all be recommended. Each has its strengths and weaknesses, and other methods are not excluded. (3) Linear regression assumes that variance across doses is constant and that the dose response is linear. If the Table 12.8

(a) In the absence of S9 mix S. typhimurium

E. coli (b) in the presence of s9 mix E. coli

a

(1) The strains are cultured overnight, and the inocula and S9 mix are prepared as in the standard Ames test. (2) The soft agar overlays are prepared and maintained at 45ºC prior to use.

Positive Controls for Use in Plate Incorporation Assays

Species

S. typhimurium

Some mutagens are poorly detected in the standard plate incorporation assay, particularly those that are metabolized to short-lived reactive electrophiles—e.g., short-chain aliphatic N-nitroso compounds (Bartsch et al., 1984). It is also possible that some metabolites may bind to components within the agar. Such compounds can be detected by using a preincubation method first described by Yahagi et al. (1975) in which the bacteria, test compound and S9 mix are incubated together in a small volume at 37ºC for a short period (30–60 min) before adding the soft agar and pouring as for the standard assay. In this variation of the test, during the preincubation step, the test compound, S9 mix and bacteria are incubated in liquid at higher concentrations than in the standard test, and this may account for the increased sensitivity with relevant mutagens. In the standard method the soluble enzymes in the S9 mix, cofactors and the test agent may diffuse into the bottom agar. This can interfere with the detection of some mutagens—a problem that is overcome in the preincubation method (Forster et al., 1980; Gatehouse and Wedd, 1984). The test is carried out as follows:

Conc. (µg plate−1)a

Strain

Mutagen

TA1535 TA100 TA1538 TA98 TA1537 WP2 uvrA

Sodium azide

1–5

Hycanthone methane sulfonate

5–20

ICR 191 Nifuroxime

1 5–15

2-Aminoanthracene

1–10

Neutral red

10–20

WP2 uvraA (pKM101) TA1538 TA1535 TA100 TA90 TA1537

The concentration given above will give relatively small increases in revertant count above the spontaneous level. There is little point in using large concentrations of reference mutagens which invariably give huge increases in revertant counts. This would give little information on the day-to-day performance of the assay.

166

Safety Evaluation of Medical Devices

(3) To each of 3–5 tubes maintained at 37ºC in a Driblock are added 0.5 mL of S9 mix, 0.1 mL of the tester strain (10–18 hr culture) and a suitable volume of the test compound, to yield the desired range of concentrations. The S9 mix is kept on ice prior to use. (4) The reaction mixtures are incubated for us to 1 hr at 37ºC. (5) 2.0 mL of soft agar is added to each tube. After mixing, the agar and reaction mixture are poured onto previously labeled, dried Vogel–Bonner plates. (6) Once the agar has set, the plates are incubated for 2–3 days before revertant colonies are scored. The use of controls is as described for the plate incorporation assay. It is crucial to use the minimum amount of organic solvent in this assay, as the total volume of the incubation mixture is small relative to the solvent component. This procedure can be modified to provide optimum conditions for particular chemical classes. For instance, preincubation times greater than 60 min plus aeration have been found necessary in the detection of allyl compounds (Neudecker and Henschler, 1985).

Forward Mutation Tests Forward mutation is an end-point that may arise from various events, including base substitutions, frameshifts, DNA deletions, etc., as mentioned earlier. Although bacterial forward mutation systems have not gained the popularity of reverse mutation tests (owing, in part, to lower sensitivity to some mutagens and lack of specificity), they have proved useful on occasion and have their supporters. Several forward mutation tests have been devised, and a brief mention of two of the more widely used systems is provided below. The ara Forward Mutation Test. The L-arabinose resistance test with Salmonella typhimurium is based on ara D mutants of the L-arabinose operon (Hera and Pueyo, 1986), ara D mutants are unable to use L-arabinose as the sole carbon source. The assay scores a change from L-arabinose sensitivity to L-arabinose resistance, which is defined as the ability to grow in a medium containing L-arabinose plus another carbon source such as glycerol. This phenotypic change reflects forward mutations in at least three different loci in the arabinose operon (Pueyo and Lopez-Barea, 1979). Strains have been constructed along the same lines as the recommended Ames strains with mutations to remove excision repair and mutations in increase permeability, and including the mutator plasmic pKM101—i.e., Salmonella typhimurium BA3 ara D531, hisG46, ∆uvrB bio and BA9 araD531, hisG46 ∆uvrB, bio, rfa (p.KM101). Protocols for the test have included plate incorporation, preincubation, and treat and plate tests (Hera and Pueyo, 1986). In the latter tests the assay does not have the problem of “plate mutants” as described for reverse mutation tests in the previous section. The recommended procedure has the following outline protocol: (1) Incubate the test strain of bacteria (107–108 cells per mL) and the test agent at 37ºC in non-selective DM medium with shaking. (2) Wash the cells after a 2 hr exposure period. (3) Plate on selective medium (DM salts, 2 mg mL−1 glycerol, 2 mL mL−1 L-arabinose, 20 µg mL−1 L-histidine, 12 µg mL−1 biotin) containing an additional supplement of D-glucose, 0.5 mg per plate.

For metabolic activation 30 µL of S9 fraction and appropriate cofactors are included in the initial incubation mixture as the standard level. Different concentrations of S9 fraction can be used as required. The group who have developed this test recommend that strain BA9 can replace the four strains used in the standard Ames test and that for the mutagens tested to date this strain detects the same range of mutagens as the Ames test strains with equal or better sensitivity. The test does seem suitable for testing complex mixtures such as red wine (Dorado et al., 1988). However, the spontaneous background count using the protocol outlined above is over 500 per plate. If fewer cells are used, false negative results are obtained (Xu et al., 1984).

Eukaryotic Mutation Tests Prokaryotic systems, as described, have proved to be quick, versatile and in many cases surprisingly accurate in identifying potential genetic hazards to man. However, there are intrinsic differences between eukaryotic and prokaryotic cells in the organization of the genome and the processing of the genetic information. Thus, there is a place for test systems based on mammalian cells for fundamental studies to understand the mutation process in higher cells and for the use of such tests for screening for genotoxic effects. The early work of Muller showed the usefulness of the fruit fly Drosophila melanogaster as a higher system for measuring germ line mutations in a whole animal. The Drosophila sex-linked recessive lethal test has yielded much useful information and in the 1970s was a popular system for screening chemicals for mutation, but this test failed to perform well in international collaborative trials to study the utility of such tests to detect carcinogens and popularity wanted. Another Drosophila test devised in the 1980s, the SMART assay (Somatic Mutation and Recombination Test) shows much promise and may revive the popularity of Drosophila for screening for genotoxic agent. There are a number of in vivo tests to measure mutation in rodents, such as the mouse specific-locus test. These are very useful for fundamental studies of radiation and chemically induced mutation, but they are rather cumbersome, are used in only a small number of expert laboratories, and are used in special circumstances where germ-line damage needs to be measured. This situation is rapidly changing as the new technologies of the last 10–15 years begin to have an impact in the construction of new model systems to measure mutation in vivo. In contrast to the situation in vivo, there are a number of test systems that use cultured mammalian cells, from both established and primary lines, that now have a large database of tested chemicals in the literature, that are relatively rapid and that are feasible to use for genetic toxicity screening. These are discussed in the next section.

In Vitro Tests for the Detection of Mammalian Mutation There have been a variety of in vitro mutation systems described in the literature, but only a small number have been defined adequately for quantitative studies (Cole et al., 1990). These are based on the detection of forward mutations in a similar manner to the systems described earlier for bacteria. A defined large number of cells are treated with the test agent and then, after a set interval, exposed to a selective toxic agent, so that only cells that have mutated can survive. As cultured mammalian cells are diploid (or near-diploid),

Chapter 12:

normally there are two copies of each gene. Recessive mutations can be missed if a normal copy is present on the homologous chromosome. As mutation frequencies for individual genes are normally very low, an impossibly large population of cells would need to be screened to detect cells in which both copies are inactivated by mutation. This problem is overcome by measuring mutation in genes on the X chromosome in male cells where only one copy of the gene will be present, or using heterozygous genes where two copies of a gene may be present but one copy is already inactive through mutation or deletion. Many genes are essential for the survival of the cell in culture, and thus mutations in such genes would be difficult to detect. However, use has been made of genes that are not essential for cell survival but allow the cell to salvage nucleotides from the surrounding medium. This saves the cell energy, as it does not have to make these compounds from simpler precursors by energy expensive catabolism. These enzymes are located at the cell membrane. If the cell is supplied with toxic nucleotides, the “normal” unmutated cells will transport these into the cell and kill the cell. However, if the cells have lost the enzyme as a result of mutation (or chromosomal deletion, rearrangement, etc.), then they will not be able to “salvage” the exogenous toxic nucleotides and will survive. The surviving mutant cells can be detected by the formation of colonies on tissue culture plates or, in some cases, in the wells of microtiter plates. One factor to take into account with these tests is that of expression time. Although a gene may be inactivated by mutation, the mRNA existing before the mutational event may decay only slowly, so that active enzyme may be present for some time after exposure to the mutagen. Thus, the cells have to be left for a period before challenging with the toxic nucleotide: this is the expression time, and differs between systems.

Chinese Hamster Lines Chinese hamster cell lines have given much valuable data over the past 15 years but their use for screening is limited by lack of sensitivity, as only a relatively small target cell population can be used, owing to metabolic co-operation (see Cole et al., 1990); however, they are still in use, so a brief description follows. Chinese hamster CHO and V79 lines have high plating efficiencies and short generation times (less than 24 hr). These properties make the lines useful for mutagenicity experiments. Both cell lines have grossly rearranged chromosomal complements, which has an unknown effect on their responsiveness to mutagens (Tweats and Gatehouse, 1988). There is some evidence that Chinese hamster lines are undergoing genetic drift in different culture collections (Kirkland, 1992).

V79 System The Chinese hamster V79 line was established in 1958 (Ford and Yerganian, 1958). Publication of the use of the line for mutation studies (by measuring resistance to purine analogues due to mutation of the Hgprt locus) occurred 10 years later (Chu and Malling, 1968). The V79 line was derived from a male Chinese hamster; hence, V79 cells possess only a single X chromosome. V79 cells grow as a cell sheet or monolayer on glass or plastic surfaces. If large numbers of cells are treated with a mutagen, when plated out, cells in close contact can link via

Genotoxicity

167

intracellular bridges. These allow the transfer of cellular components between cells such as messenger RNA. Thus, if a cell carries a mutation in the hgprt gene resulting in the inactivation of the relevant mRNA, it can receive viable mRNA or intact enzyme from adjacent non-mutated cells. Therefore, when the mutated cell is challenged with a toxic purine, it is lost, owing to the presence of active enzyme derived from the imported mRNA. This phenomenon is termed “metabolic co-operation” and severely limited the sensitivity of lines such as V79 for mutagen detection. This drawback can be overcome to an extent by carrying out the detection of mutant clones in semi-solid agar (see, e.g., Oberly et al., 1987) or by using the “respreading technique” (see, e.g., Fox, 1981). The preferred expression time for Hgprt mutants is 6–8 days, although care needs to be taken when testing chemicals well into the toxic range, where the “expression time” needs to be extended to allow recovery.

Preliminary Cytotoxicity Testing An essential first step is to carry out a preliminary study to evaluate the toxicity of the test material to the indicator cells, under the conditions of the main mutagenicity test. When selecting dose levels, the solubility of the test compound, the resulting pH of the media and the osmolality of the test solutions all need to be considered. The latter two parameters have been known to induce false positive effects in in vitro mammalian tests (see, e.g., Brusick, 1986). The experimental procedure is carried out as follows: (1) Seek T75 plastic tissue culture flasks with a minimum of 2.5 × 106 cells in 120 mL of Eagle’s medium containing 20 mML-glutamine: 0.88 g L−1 sodium bicarbonate; 20 mM HEPES; 50 µg mL−1 streptomycin sulphate; 50 IU mL−1 benzylpenicillin; and 7.5% of fetal bovine serum. The flasks are incubated for 18–24 hr at 37ºC in a CO2 incubator to establish monolayer cultures. (2) Prepare treatment medium containing various concentrations of test compound—e.g., 19.7 mL of Eagle’s medium (without serum) plus 300 µL of stock concentration of compound in a preferred solvent (e.g., water, ethanol, DMSO, etc.). The final concentration of solvent other than water should not exceed 1% v/v/. Normally a range of 0–5000 µg mL−1 (final concentration) is covered. For a sparingly soluble compound, the highest concentration will be the lowest at which visible precipitation occurs. Similarly, if a compound has a marked effect on osmolality, concentrations should not be used that exceed 500 milliosmoles (mosm) per kg. In addition, a pH range of 6.5–7.5 should not be exceeded. (3) Each cell monolayer is rinsed with a minimum of 20 mL phosphate buffered saline (PBS) and then 20 mL of treatment medium is carefully added. The flasks are incubated for 3 hr at 37ºC in a CO2 incubator. (4) After treatment, carefully discard the medium from each flask and wash each monolayer twice with PBS. Care needs to be taken safely to dispose of contaminated solutions. (5) 10 mL of trypsin solution (0.025% trypsin in PBS) is added to each flask. Once the cells have rounded up, the trypsin is neutralized by the addition of 10 mL of complete medium. A cell suspension is obtained by vigorous pipetting to break up cell clumps. (6) The trypsinized cell suspension is counted and diluted in complete media before assessing for survival. For each treatment set up five Petri dishes containing 200 cells per dish.

168

Safety Evaluation of Medical Devices

(7) Incubate at 37ºC in a CO2 incubator for 7–10 days. (8) The medium is removed and the colonies are fixed and stained, using 5% Giemsa in buffered formalin. Once the colonies are stained, the Giemsa is removed and the colonies are counted. The method can be repeated including 20% v/v S9 mix. To calculate percentage survival, the following formula is used: Mean no. of colonies on treated plates Cell titre in treated culture × × 100 Cell titre in control culture Mean no. of colonies on control plates The cloning efficiency (CE) of the control culture is calculated as follows: CE =

Mean no. of colonies per plate × 100 No. of cells per plate (i.e., 200)

In the absence of precipitation or effects on pH or osmolality, the maximum concentration of the main mutagenicity study is a concentration that reduces survival to approximately 20% of the control value.

Procedure for the Chinese Hamster V79/Hgprt Assay The assay usually comprises three test concentrations, each in duplicate, and four vehicle control replicates. Suitable positive controls are ethylmethane sulfonate (−S9) and dimethyl benzanthracene (+S9). V79 cells with a low nominal passage number should be used from frozen stocks to help minimize genetic drift. The procedure described includes a reseeding step for mutation expression. Steps 1–5 are the same as the cytotoxicity assay. As before, tests can be carried out in the presence and in the absence of S9 mix. (6) The trypsinized cultures are counted and a sample is assessed for survival as for the cytotoxicity assay. In addition, an appropriate number of cells are reseeded for estimation of mutation frequency at the day 8 expression time. The cells are transferred to roller bottles (usually 490 cm2) for this stage. The bottles are gassed with pure CO2, the tops are tightened and the bottles are incubated at 37ºC on a roller machine (approximate speed 0.5–1.0 rev min−1). Usually 106 viable cells are reseeded in 50 mL of Eagle’s medium containing serum, but more cells are required at the toxic dose levels. (7) The bottles are subcultured as necessary throughout the expression period to maintain sub-confluency. This involves retrypsinization and determining the cell titre for each treatment. For each culture a fresh roller bottle is reseeded with a minimum of 106 cells. (8) On day 8, each culture is again trypsinized, counted and diluted so that a sample cell population can be assessed for cloning efficiency and a second sample can be assessed for the induction of 6TG-resistant cells. (9) The cell suspension is diluted in complete medium and 2 × 105 cells added per Petri dish (10 Petri dishes per treatment). 6-Thioguanine is added to the medium at a final concentration of 10 µg mL−1. (10) The Petri dishes are incubated for 7–10 days and the medium is then removed. The colonies are fixed and

stained as previously. The colonies (>50 cells per clone) are then counted. Mutation frequency in each culture is calculated as: Mean no. colonies on thioguanine plates 1000 × Mean no. colonies on survival plates

Data Analysis (Arlett et al., 1989) A weighted analysis of variance is performed on the mutation frequencies, as the variation in the number of mutations per plate usually increases as the mean increases. Each dose of test compound is compared with the corresponding vehicle control by means of a one-sided Dunnett’s test and, in addition, the mutation frequencies are examined to see whether there is a linear relationship with dose. The criterion employed for a positive response in this assay is a reproducible statistically significant increase in mutation frequency (weighted mean for duplicate treated cultures) over the concurrent vehicle control value (weighted mean for four independent control cultures). Ideally, the response should show evidence of a dose–response relationship. When a small isolated significant increase in mutation frequency is observed in only one of the two duplicate experiments, then a third test should be carried out. If the third test shows no significant effects, the initial increase is likely to be a chance result. In cases where an apparent treated-related increase is thought to be a result of unusually low variability or a low control frequency, comparison with the laboratory historical control frequency may be justified.

Chinese Hamster CHO/Hgprt System Chinese hamster ovary (CHO) cells have 21 or 22 chromosomes with one intact X chromosome and a large acrocentric marker chromosome (Natarajan and Obe, 1982). The use of these cells in mammalian mutation experiments was first reported by Hsie et al. (1975), and was refined into a quantitative assay for mutagenicity testing by O’Neill. The performance of this system has been reviewed by the USA EPA Gene-Tox Program. The experimental procedure for this assay is similar to the V70/Hgprt system already described, and for more detailed descriptions the reader is referred to Li et al. (1987).

Mouse Lymphoma L5178Y TK+/− Assay Whereas the Chinese hamster cell systems are regarded as relatively insensitive, the mouse lymphoma L5178Y TK+/− test is undoubtedly more sensitive. Unfortunately, there are persistent doubts regarding its specificity—i.e., the ability to distinguish between carcinogens and non-carcinogens (e.g., Tennant et al., 1987). However, a great advantage is the ability of these cells to grow in suspension culture in which intracellular bridges do not occur. Thus, the problems of metabolic co-operation are avoided, which allows a large number of cells to be treated for optimum statistical analysis of results. A candid historical overview of the development of the mouse lymphoma TK+/− mutagenicity assay is given by its originator, Clive (1987). Initially methodologies were developed for producing the three TK genotypes (TK+/+ and TK−/− homozygotes and the TK+/− heterozygotes (Clive et al., 1972). This first heterozygote was lost; however, it was recognized that subsequent heterozygotes produced distinctly bimodal distributions of mutant-colony sizes, owing to

Chapter 12:

differences in growth rate. These were interpreted in terms of single-gene (large-colony mutants) and viable chromosomal mutations (small-colony mutants). A period of diversification of the mouse lymphoma assay followed and controversy over the significance of small-colony mutants (e.g., Amacher et al., 1980). Following this, a series of cytogenetic studies confirmed the cytogenetic interpretation for small-colony mutants (e.g., Hozier et al., 1982). Molecular studies showed that most mutations resulting in small-colony mutants involve largescale deletions (Evans et al., 1986). A current theory states that, for many compounds, deletion mutants are induced by binding of the compound to complexes between topoisomerase II and DNA (Clive, 1989). Topoisomerases are enzymes that control supercoiling via breakage and reunion of DNA strands; it is the latter strep that is disrupted, which leads to chromosome damage and deletions. Further molecular studies (Applegate et al., 1990) have shown that a wide variety of genetic events can result in the formation of TK+/− genotype from the heterozygote, including recombinations and mitotic non-disjunction. The TK+/− line was originally isolated as a spontaneously arising revertant clone from a UV-induced TK−/− clone. The parental TK+/+ cell and the heterozygote were then the only TK-competent mouse lymphoma cells that could be maintained in THMG medium (3 µg mL−1 thymidine, 5 µg mL−1 hypoxanthine, 0.1 µg mL−1 methotrexate, and 7.5 µg mL−1 glycine) (Clive, 1987). Thus, like most established lines, these cells are remote from wild-type cells. The karyotype of the TK+/− 3.7.2 C line has a modal chromosome number of 40 like wild-type, but has a variety of chromosomal rearrangements and centromeric heteromorphisms (Blazak et al., 1986). Two main protocols have been devised for carrying out mutation assays with mouse lymphoma L5178Y cells— i.e., plating the cells in soft agar or a fluctuation test approach. It is the latter that is described in the following section, based on Cole et al. (1986). The reader is referred to Clive et al. (1987) for a full description of the soft-agar method.

Preliminary Cytotoxicity Assay The cells are maintained in RPMI 1640 medium containing 2.0 mM glutamine, 20 mM HEPES, 200 µg mL−1 sodium pyruvate, 50 IU mL−1 benzylpenicillin, 50 µg mL−1 streptomycin sulfate and 10% donor horse serum (heat-inactivated for 30 min at 56ºC). This medium is designated CM10. Conditioned medium is CM10 in which cells have grown exponentially for at least 1 day. Treatment medium contains 3% horse serum and 30% conditioned media (CM3). Medium without serum is known as incomplete medium (ICM). If treatment time exceeds 3 hr, treatment is carried out in CM10. The method is as follows: (1) The cell titre of an exponentially growing culture of cells in CM10 is determined with a Coulter counter. The cell suspension is centrifuged at 70 g for 5 min and the supernatant is reduced such that 3 mL contains approximately 5 × 106 cells (3 hr treatment) or 2 × 106 (treatment >3 hr). (2a) For tests in the absence of S9 mix, treatment groups are prepared by mixing 3 mL of solution of test compound and 6.9 mL of ICM (3 hr treatment) or 6.9 mL of CM10 (treatment >3 hr).

Genotoxicity

169

(2b) Tests in the presence of S9 mix are carried out in the same way, except the treatment medium contains 10% v/v S9 mix at the expense of ICM—i.e., 3 mL cell suspension, 5.9 mL ICM, 1 mL S9 mix and 0.1 mL test compound solution/vehicle. The composition of the S9 mix is as described earlier (p. 896). It is prepared immediately before required and kept on ice until it is added to the test system. For the vehicle controls, if an organic solvent is used, it should not exceed 1% v/v. (3) After the treatment period, cells are spun down at 70 g for 5 min and the supernatant is transferred for assessment of pH (pH meter) and osmolality (e.g., using Wescor Vapour Pressure Osmometer). The cell pellet is washed twice in PBS and then resuspended in 10 mL CM10. (All contaminated material and waste should be disposed of safely). (4) The cell titre of each culture is counted and a sample diluted in CM10 for assessment of post-treatment survival. For this two 96-well microtitre plates are charged with 200 µL of a diluted cell suspension, using a multi-channel pipette such that each well contains on average one cell. (5) Plates are incubated for 7–8 days at 37ºC and 5% CO2 in 95 ± 3% relative humidity. (6) The plates are removed from the incubator and 20 µl of MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] at 5 mg ml−1 (in PBS) is added to each well with a multichannel pipette. The plates are left to stand for 1-4 h and are then scored for the presence of colonies with a Titertek mirror-box, which allows direct viewing of the bottom surface of the plates. (7) Cytotoxicity can also be determined post-treatment as follows: T25 flasks are set up after treatment containing 0.75 × 105 cells per ml in 5 ml CM10. Flasks are incubated with loose lids at 37ºC with 5 per cent CO2 in 95 ± 3 per cent relative humidity. Two days later the cell titre of each culture is determined with a Coulter counter. (8) Following this procedure, various calculations are carried out to aid selection of dose levels for the main mutation assay. (a) Cloning efficiency. In microtitre assays calculations are based on the Poisson distribution—i.e.,:

)

P (o =

No. of wells without a colony Total no. of cells

(b) Relative survival: Relative survival (S) is calculated as follows: S=

CE of treated group CE of control group

(c) Growth: Growth in suspension (SG) is calculated as follows: SG =

Cell count after 3 days 0.75 × 10 5

Relative suspension growth (RSG) is calculated as follows: RSG =

SG of treated group × 100% SG of control group

170

Safety Evaluation of Medical Devices

Selection of Dose Levels The highest test concentration is selected from one of the following options, whichever is lowest: • A concentration which reduces survival to about 10–20% of the control value. • A concentration which reduces RSG to 10–20% of the control value. • The lowest concentration at which visible precipitation occurs. • The highest concentration which does not increase the osmolality of the medium to greater than 400 mmol kg−1 or 100 mmol above the value for the solvent control. • The highest concentration that does not alter the pH of the treatment medium beyond the range 6.8–7.5. • If none of these conditions are met, 5 mg ml−1 should be used. Lower test concentrations are selected as fractions of the highest concentration, usually including one dose which causes 20–70% survival and one dose which causes >70% survival.

Main Mutation Assay The assay normally comprises three test concentrations, a positive control and vehicle control. All treatment groups are set up in duplicate. The expression time is 2 days, unless there are indications that the test agent inhibits cell proliferation, where an additional or possibly alternative expression time should be employed. Stock cultures are established from frozen ampoules of cells that have been treated with thymidine, hypoxanthine, methotrexate and glycine for 24 hr, which purges the culture of pre-existing TK−/− mutants. This cell stock is used for a maximum of 2 months. Treatment is normally carried out in 50 mL centrifuge tubes on a roller machine. During the expression time the cells are grown in T75 plastic tissue culture flasks. For estimation of cloning efficiency and mutant induction, cells are plated out in 96-well microtitre plates. Flasks and microtitre plates are incubated at 37ºC in a CO2 incubator as in the cytotoxicity assays. Cell titres are determined by diluting of the cell suspension in Isoton and counting an appropriate volume (usually 0.5 mL) with a Coulter counter. Two counts are made per suspension. The experimental procedure is carried out as follows: (1) On the day of treatment stock solutions for the positive control and the various concentrations of test compound (selected as per the previous selection) are prepared. (2) Treatment is carried out in 30% conditioned media. The serum concentration is 3% (3 h treatment) or 10% (treated >3 h). (3) Cell suspensions of exponentially growing cells are prepared as in the cytotoxicity assay, except that 6 mL of media required for each treatment culture contains 107 cells (3 h treatment) or 3 × 106 cells (>3 h treatment). The number of cells per treatment may be increased if marked cytotoxicity is expected, to allow enough cells to survive (e.g., if 20% survival or less is expected, 2 × 107 cells may be treated). (4) For tests in the absence of S9 mix, 6 mL of cell suspension, 0.2 mL test compound/vehicle and 13.8 mL ICM (3 h treatment) or 13.8 mL CM10 (treatment >7 h) are

mixed in the presence of S9 mix and 0.2 mL of test compound/vehicle are prepared. (5) After treatment the cells are centrifuged at 70 g for 5 min, and supernatant is discarded and the cell pellet is resuspended in PBS (pH 7). This washing procedure is repeated twice, and finally the cell pellet is resuspended in CM10. (6) Each culture is counted so that a sample of cells can be assessed for post-treatment survival, and the remaining cell population assessed for estimation of mutation frequency. (7) For survival estimation, cells are placed into 96-well microtitre trays at a cell density of 1 cell per well as per the cytotoxicity assay. (8) For mutation estimation, the cells are diluted to a cell density of 2 × 105 cells per mL with CM10 in tissue culture flasks and the culture is incubated at 37ºC in a CO2 incubator. On day 1 each culture is counted and diluted with fresh medium to a cell density of 2 × 105 cells per mL in a maximum of 100 mL of medium. (9) On day 2 each culture is counted again and an aliquot of cells taken so that: (a) a sample of the cell population can be assessed for cloning efficiency. Plates are incubated at 37ºC in a CO2 incubator for 7 days; (b) a sample of the cell population can be assessed for the induction of TFTresistant cells (mutants). For this 2 × 103 cells are plated per well in 200 µL CM10 containing 4 µg mL−1 TFT. TFT and TFT-containing cultures must not be exposed to bright light, as the material is light-sensitive. The plates are incubated for 10–12 days at 37ºC in a CO2 incubator. (10) At the end of incubation 20 µL MTT is added to each well. The plates are left to develop for 1–4 h at 37ºC and then scored for colony-bearing wells. Colonies are scored by eye and are classified as small or large. The calculation for cloning efficiency is made as for the cytotoxicity assay. Relative total growth (RTG) is a cytotoxicity parameter which considers growth in suspension during the expression time and the cloning efficiency of the end of the expression time as follows: Suspension growth (SG) = RTG =

24 hr cell count 48 hr cell count × 2 × 10 5 2 × 10 4

SG treated culture CE of treated culture × SG control culture CE of control culture

Mutation frequency (MF) is calculated as follows: M=

InP0 for mutation plates No. of cells per well × CE/100

Data Analysis Data from the fluctuation test described above are analyzed by an appropriate statistical method as described in Robinson et al. (1989). Data from plate assays are analyzed as described in Arlett et al. (1989) for treat and plate tests.

Status of Mammalian Mutation Tests At present the only practical assays for screening new chemical entities for mammalian mutation are the mammalian cell assays described above. The protocols are well-defined, and mutant selection and counting procedures

Chapter 12:

are simple and easily quantified. In general, the genetic end-points are understood and relevant to deleterious genetic events in humans. For these reasons the assays are still regarded as valuable in safety evaluation (Li et al., 1991). It is, however, recognized that there are still unknown factors and molecular events that influence test results. This can be illustrated by the conclusions of the third UKEMS collaborative trial, which focused on tests with cultured mammalian cells. The following points were made: • The number of cells to be cultured during expression imposes a severe limitation in the use of surface attached cells. • The importance of a careful determination of toxicity. • That S9 levels may need to be varied. • That the aromatic amine benzidine is mutagenic only at the TK locus in L5178Y TK+/− cells. The most disturbing finding was that benzidine (detectable without metabolism by S9 mix) did not produce detectable DNA adducts (as shown by 32P-post-labelling) in L5178Y cells. Thus, the mechanism for mutagenesis in L5178Y cells benzidine remains to be elucidated (Arlett and Cole, 1990).

In Vivo Mammalian Mutation Tests Mammalian mutation studies of chemicals in the whole animal have provided fundamental information on mutation parameters in germ cells such as dose response, dose fractionation, sensitivity of various stages in gametogenesis, etc., just as is known for ionizing radiation (Russell, 1989). This has led to estimations of the possible impact chemical mutagens may have on heritable malformation, inborn errors of metabolism, etc. Today germ cell studies are still required when estimating the heritable damage a mutagen may inflict on exposed human populations. The existing tests tend to be cumbersome and are not used for routine genetic toxicology screening, and thus only brief descriptions will follow. Reviews of existing data, particularly by Holden (see, e.g., Holden, 1982; Adler and Ashby, 1989), have indicated that most if not all germ cell mutagens also induce DNA damage in somatic cells, as detected by well-established assays such as the rodent micronucleus test. The converse is not true i.e., some mutagens/clastogens can induce somatic cell damage but do not induce germ cell changes, which probably reflects the special protection afforded to the germ cells, such as that provided by the blood–testis barrier. In other words, it appears that germ cell mutagens are a subset of somatic cell mutagens. In vivo mammalian mutation tests are not restricted to germ cell tests. The mouse spot test described below is, again, a test used first for studying radiation-induced mutation but has also been used for screening chemicals for in vivo mutagenic potential. This test has had several proponents but compared with in vivo chromosomal assays is not widely used.

The Mouse Specific Locus Test The mouse somatic spot test is a type of specific locus test. The classical specific locus test was developed independently by Russell at Oak Ridge in the late 1940s (Russell, 1951, 1989) and Carter in Edinburgh (Carter et al., 1956). The test consists of treatment of parental mice homozygous for a wild-type set of marker loci. The targets for mutation are the germ cells in the gonads of the treated mice. These are mated with a tester stock that is homozygous recessive at the marker loci. The F1 offspring that result are normally

Genotoxicity

171

heterozygous at the marker loci and thus express the wild-type phenotype. In the event of a mutation from the wild-type allele at any of these loci, the F1 offspring express the recessive phenotype. The test marker strain (T) developed by Russell uses seven recessive loci, viz: a (non-agouti), b (brown), cch (chinchilla), d (dilute), p (pink-eyed dilution), s (piebald), and se (short-ear). As for the mouse spot test, these genes control coat pigmentation, intensity or pattern, and, for the se gene, the size of the external ear. As the occurrence of mutation is rare even after mutagen treatment, the specific locus test is the ultimate study of mutation, requiring many thousands of offspring to be scored, plus significant resources of time, space, and animal husbandry. Because of these constraints it is often difficult to define a negative result, as insufficient animals are scored or all stages of spermatogenesis are not covered. Of the 25 compounds tested in the assay, as reviewed by Ehling et al. (1986), 17 were regarded as “inconclusive” and 8 positive. The scale studies can reach is illustrated by the test of ethylene oxide described by Russell et al. (1984), where exposures of 101,000 and 150,000 ppm per hour were used over 16–23 weeks. A total of 71,387 offspring were examined. The spermatogonial stem-cell mutation rate in the treated animals did not differ significantly from the historical control frequency! With regard to the design of the test, mice are mated when 7–8 weeks old. By this age all germ cell stages are present. The test compound is normally administered by the ip route to maximize the likelihood of germ cell exposure. The preferred dose is just below the toxic level so long as fertility is not compromised. One lower dose should also be included. In males spermatogonia are most at risk but it is also desirable that later stages also be exposed. Thus, the mice are mated immediately after treatment to 2–4 females. This is continued each week for 7 weeks. Then the first group has completed its rearing of the first set of offspring and is remated. This cycle can be continued for the lifetime of the males. Tests can also be carried out by dosing females, when treatment is carried out for 3 weeks to cover all stages of ogenesis. The offspring are examined immediately after birth for identification of malformations (dominant visibles) and then at weaning for the specific locus mutations. Presumptive mutant mice are checked by further crosses to confirm their status (Searle, 1984). Comparison of mutation frequencies is made with the historical database. For definition of a positive result the same principles are recommended as for the mouse spot test (Selby and Olson, 1981). A minimum size of 18,000 offspring per group is recommended by those authors for definition of a negative result.

IN VITRO CYTOGENETIC ASSAYS The in vitro cytogenetic assay is a short-term mutagenicity test for detecting chromosomal damage in cultured mammalian cells. Cultured cells have a limited ability metabolically to activate some potential clastogens. This can be overcome by adding an exogenous metabolic activation system such as S9 mix to the cells (Ames et al., 1975; Natarajan et al., 1976; Maron and Ames, 1983; Madle and Obe, 1980).

172

Safety Evaluation of Medical Devices

Observations are made in metaphase cells arrested with a spindle inhibitor such as colchicine or colcemid to accumulate cells in a metaphase-like stage of mitosis (c-metaphase) before hypotonic treatment to enlarge cells and fixation with alcohol/acetic acid solution. Cells are then dispersed on to microscope slides and stained and slides are randomized, coded and analyzed for chromosome aberrations with high-power light microscopy. Details of the procedure are given in Dean and Danford (1984) and Preston et al. (1981, 1987). The UKEMS guidelines (Scott et al., 1990) recommend that all tests be repeated regardless of the outcome of the first test and that, if a negative or equivocal result is obtained in the first test, the repeat should include an additional sampling time. In the earlier version of the guidelines (Scott et al., 1983) a single sampling at approximately 1.5 normal cycle times (−24 hr for a 1.5 cell cycle) from the beginning of treatment was recommended, provided that a range of concentrations was used which induced marginal to substantial reductions in mitotic index, usually an indicator of mitotic delay. However, Ishidate (1988a) reported a number of chemicals which gave negative responses with a fixation time of 24 hr but which were positive at 48 hr. This was when a Chinese hamster fibroblast line (CHO) with a doubling time of 15 hr was used. It would appear, therefore, that there are chemicals which can induce extensive mitotic delay at clastogenic doses and may be clastogenic only when cells have passed through more than one cell cycle since treatment (Thust et al., 1980). A repeat test should include an additional sample at approximately 24 hr later but it may only be necessary to score cells from the highest dose at this later fixation time. When the first test gives a clearly positive result, the repeat test need only utilize the same fixation time. The use of other sampling times is in agreement with other guidelines (European Community EEC Directive— OECD, 1983; American Society for Testing and Materials— Preston et al., 1987; Japanese Guidelines—JMHW, 1984; Joint Directives, 1987; Ishidate, 1988b).

Cell Types Established cell lines, cell strains or primary cell cultures may be used. The most often used are Chinese hamster cell lines and human peripheral blood lymphocytes. The merits of these two cell lines have been reported (Ishidate and Harnois, 1987; Kirkland and Garner, 1987). The cell system must be validated and consistently sensitive to known clastogens.

Chinese Hamster Cell Lines Chinese hamster cell lines have a small number of large chromosomes (11 pairs). Chinese hamster ovary cells in which there has been an extensive rearrangement of chromosome material and the chromosome number may not be constant from cell to cell, are frequently used. Polyploidy, endoreduplication and high spontaneous chromosome aberration frequencies can sometimes be found in these established cell lines, but careful cell culture techniques should minimize such effects. Cells should be treated in exponential growth when cells are in all stages of the cell cycle.

Human Peripheral Blood Lymphocytes Blood should be taken from healthy donors not known to be suffering from viral infections or receiving medication. Staff handling blood should be immunized against hepatitis B and regular donors should be shown to be hepatitis B antigen negative. Donors and staff should be aware of AIDS implications, and blood and cultures should be handled at

containment level 2 (Advisory Committee on Dangerous Pathogens, 1984). Peripheral blood cultures are stimulated to divide by the addition of a T cell mitogen such as phytohaemagglutinin (PHA) to the culture medium. Mitotic activity is at a maximum at about 3 days but begins at about 40 hr after PHA stimulation and the chromosome constitution remains diploid during short-term culture (Evans and O’Riordan, 1975). Treatments should commence at about 44 hr after cultre initiation. This is when cells are actively proliferating and cells are in all stages of the cell cycle. They should be sampled about 20 hr later. In a repeat study the second sample time should be about 92 hr after culture initiation. Morimoto et al. (1983) report that the cycle time for lymphocytes averages about 12–14 hr except for the first cycle. Female donors can give higher yields of chromosome damage (Anderson et al., 1990).

Positive and Negative Controls When the solvent is not the culture medium or water, the solvent, liver enzyme activation mixture, and solvent and untreated controls are used as negative controls. Since cultured cells are normally treated in their usual growth medium, the solubility of the test material in the medium should be ascertained before testing. Extremes of pH can be clastogenic (Cifone et al., 1987), so the effect of the test material on pH should also be determined, but buffers can be utilized. Various organic solvents are used, such as dimethyl sulfoxide (DMSO), dimethylformamide, ethanol, and acetone. The volume added must not be toxic to cells. Greater than 10% water v/v can be toxic because of nutrient dilution and osmolality changes. A known clastogen should always be included as a positive control. When metabolic activation is used, a positive control chemical known to require metabolic activation should also be used to ensure that the system is functioning properly. Without metabolic activation, a direct-acting positive control chemical should be used. A structurally related positive control can also be used. Appropriate safety precautions must be taken in handling clastogens (IARC, 1979; MRC, 1981). Positive control chemicals should be used to produce relatively low frequencies of aberrations so that the sensitivity of the assay for detecting weak clastogens can be established (Preston et al., 1987). Aberration yields in negative and positive controls should be used to provide a historical database.

Treatment of Cells When an exogenous activation system is employed, short treatments (about 2 hr) are usually necessary because S9 mix is often cytotoxic when used for extended lengths of time. However, cells may be treated with chemicals either continuously up to harvest time or for a short time followed by washing and addition of fresh medium to allow cell cycle progression. Continuous treatment avoids centrifugation steps required with washing of cells and optimizes the endogenous metabolic capacity of the lymphocytes. When metabolic activation is used, S9 mix should not exceed 1–10% of the culture medium by volume. It has been shown that the S9 mix is clastogenic in CHO cells and mouse lymphoma cells (Cifone et al., 1987; Kirkland et al., 1987) but not in human lymphocytes, where blood components can inactivate active oxygen species which could cause chromosome damage. When S9 mix from animals treated with

Chapter 12:

other enzyme-inducing agents such as phenobarbitone/ beta-naphthoflavone, is used, clastogenesis may be minimized (Kirkland et al., 1989). Prior to testing, it is necessary to determine the cytotoxicity of the test material, in order to select a suitable dose range for the chromosome assay both with and without metabolic activation. The range most commonly used determines the effect of the agent on the mitotic index (MI), i.e., the percentage of cells in mitoses at the time of cell harvest. The highest dose should inhibit mitotic activity by approximately 50% (EEC Annex V), 75% (UKEMS: Scott et al., 1990) or exhibit some other indication of cytotoxicity. If the reduction in MI is too great, insufficient cells can be found for chromosome analysis. Cytotoxicity can also be assessed by making cell counts in the chromosome aberration test when using cell lines. In the lymphocyte assay total white cell counts can be used in addition to MI. A dose which induces 50–75% toxicity in these assays should be accompanied by a suitable reduction in mitotic index. If the test material is not toxic, it is recommended by, for example, the EEC (Annex V) that it be tested up to 5 mg mL−1. The UKEMS recommends that chemicals be tested up to their maximum solubility in the treatment medium and not just their maximum solubility in stock solutions. For highly soluble non-toxic agents, concentrations above 10 mM may produce substantial increases in the osmolality of the culture medium which could be clastogenic by causing ionic imbalance within the cells (Ishidate et al., 1984; Brusick, 1987). At concentrations exceeding 10 mM the osmolality of the treatment media should be measured and if the increase exceeds 50 mmol kg−1, clastogenicity resulting from high osmolality should be suspected and, according to the UKEMS, is unlikely to be of relevance to human risk. The UKEMS also does not recommend the testing of chemicals at concentrations exceeding their solubility limits as suspensions or precipitate. A minimum of three doses of the test material should be used—the highest chosen as described above, the lowest on the borderline of toxicity and an intermediate one. Up to six doses can be managed satisfactorily, and this ensures the detection of any dose response and that a toxic range is covered. MIs are as required for the preliminary study (at lease 1000 cells per culture). It is also useful to score endoreduplication and polyploidy for historical data. Cells from only three doses need to be analyzed. The range of doses used at the repeat fixation time can be those which induce a suitable degree of mitotic inhibition at the earlier fixation time, but if the highest dose reduces the MI to an unacceptably low level at the second sampling time, the next highest dose should be chosen for screening. A complete assay requires the test material to be investigated at a minimum of three doses together with a positive (untreated) and solvent-only control can be omitted if tissue culture medium is used as a solvent. When two fixation times are used in repeat tests, the positive control is necessary at only one time but the negative or solvent control is necessary at both times. Duplicates of each test group and quadruplicates of solvent or negative controls should be set up. The sensitivity of the assay is improved with larger numbers scored in the negative controls (Richardson et al., 1989).

Scoring Procedures Prior to scoring, slides should be coded, randomized and then scored “blind.” Metaphase analysis should only be

Genotoxicity

173

carried out by an experienced observer. Metaphase cells should be sought under low-power magnification and those with well-spread, i.e., non-overlapping, clearly defined non-fuzzy chromosomes examined under high power with oil immersion. It is acceptable to analyze cells with total chromosome numbers or that have lost one or two chromosomes during processing. In human lymphocytes (2n-46) 44 or more centomeres and in CHO cells (2n = 22; range 21–24) 20 or more centromeres can be scored. Chromosome numbers can be recorded for each cell, to give an indication of aneuploidy. Only cells with increases in numbers (above 46 in human lymphocytes and 24 in CHO cells) should be considered in this category, since decreases can occur through processing. Recording microscope co-ordinates of cells is necessary and allows verification of abnormal cells. A photographic record is also useful of cells with aberrations. Two hundred cells (100 from each of two replicates) should be scored per treatment group. When ambiguous results are obtained, there may be further “blind” reading of these samples.

Data Recording The classification and nomenclature of the International System for Human Cytogenetic Nomenclature (ISCN, 1985) as applied to acquired chromosome aberrations is recommended. Score sheets giving the slide code, microscope scorer’s name, date, cell number, number of chromosomes, and aberration types should be used. These should include chromatid and chromosome gaps, deletions, exchanges, and others. A space for the Vernier reading for comments and a diagram of the aberration should be available. From the score sheets, the frequencies of various aberrations should be calculated and each aberration should be counted only once. To consider a break as one event and an exchange as two events is not acceptable, since unfounded assumptions are made about mechanisms involved (Revell, 1974).

Presentation of Results The test material, test cells used, method of treatment, harvesting of cells, cytotoxicity assay, etc., should be clearly stated as well as the statistical methods used. Richardson et al. (1989) recommend that comparison be made between the frequencies in control cells and at each dose level using Fisher’s Exact Test. In cytogenetic assays the absence of a clear positive dose–response relationship at a particular time frequently arises. This is because a single common sampling time may be used for all doses of a test compound. Chromosome aberration yields can vary markedly with post-treatment sampling time of an asynchronous population, and increasing doses of clastogens can induce increasing degrees of mitotic delay (Scott et al., 1990). Additional fixation times should clarify the relationship between dose and aberration yield. Gaps are by tradition excluded from quantification of chromosome aberration yields. Some gaps have been shown to be real discontinuities in DNA (e.g., Heddle and Bodycote, 1970). Where chromosome aberration yields are on the borderline of statistical significance above control values, the inclusion of gaps could be useful. Further details on this approach may be found in the UKEMS guidelines (Scott et al., 1990). Since chromosome exchanges are relatively rare events, greater biological significance should be attached to their presence than to gaps and breaks.

174

Safety Evaluation of Medical Devices

1

2

3

6

7

8

13

14

15

19

20

Chemicals which are clastogenic in vitro at low doses are more likely to be clastogenic in vivo than those where clastogenicity is detected only at high concentrations (Ishidate et al., 1988). Negative results in well-conducted in vitro tests are a good indication of a lack of potential for in vivo clastogenesis, since almost all in vivo clastogens have given positive results in vitro when adequately tested (Thompson, 1986; Ishidate et al., 1988).

IN VIVO CYTOGENETICS ASSAYS Damage induced in whole animals can be detected in in vivo chromosome assays in either somatic or germinal cells by examination of metaphases or the formation of micronuclei. The micronucleus test can also detect whole chromosome loss or aneuploidy in the absence of clastogenic activity and is considered comparable in sensitivity to chromosome analysis (Tsuchimoto and Matter, 1979). Rats and mice are generally used for in vivo studies, with the mouse being employed for bone marrow micronucleus analysis and the rat for metaphase analysis, but both can be used for either. Mice are cheaper and easier to handle than rats, and only a qualitative difference in response has been found between the species (Albanese et al., 1988). Chinese hamsters are also widely used for metaphase analysis because of their low diploid chromosome number of 22. However, there are few other historical toxicological data for this species.

Somatic Cell Assays Metaphase Analysis Metaphase analysis can be performed in any tissue with actively dividing cells, but bone marrow is the tissue most often examined. Cells are treated with a test compound and are arrested in metaphase by the administration of colcemid or colchicine at various sampling times after treatment.

9

4

5

10

11

12

16

17

18

21

22

X

Y

Preparations are examined for structural chromosome damage. Because the bone marrow has a good blood supply, the cells should be exposed to the test compound or its metabolites in the peripheral blood supply, and the cells are sensitive to S-dependent and S-independent mutagens (Topham et al., 1983). Peripheral blood cells can be stimulated to divide even though the target cell is relatively insensitive (Newton and Lilly, 1986). It is necessary to stimulate them with a mitogen since the number of lymphocytes which are dividing at any one time is very low. Cells are in G0 when exposure is taking place, so they may not be sensitive to cell cycle stage specific mutagens and any damage might be repaired before sampling.

Micronuclei The assessment of micronuclei is considered simpler than the assessment of metaphase analysis. This assay is most often carried out in bone marrow cells, where polychromatic erythrocytes are examined. Damage is induced in the immature erythroblast and results in a micronucleus outside the main nucleus, which is easily detected after staining as a chromatid-containing body. When the erythroblast matures, the micronucleus, whose formation results from chromosome loss during cell division or from chromosome breakage forming centric and acentric fragments, is not extruded with the nucleus. Micronuclei can also be detected in peripheral blood cells (MacGregor et al., 1980). In addition, they can be detected in liver (Tates et al., 1980; Braithwaite and Ashby, 1988) after partial hepatectomy or stimulation with 4-acetylaminofluorene, or they can be detected in any proliferating cells.

Germ Cell Assays The study of chromosome damage is highly relevant to the assessment of heritable cytogenetic damage. Many compounds which cause somatic cell damage have not produced germ cell damage (Holden, 1982)

Chapter 12:

and, so far, all germ mutagens have also produced somatic damage. Germ cell data, however, are needed for genetic risk estimation, and testing can be performed in male or female germ cells. The former are most often used, owing to the systemic effects in females. Testing in the male is performed in mitotically proliferating premeiotic spermatogonia, but chromosomal errors in such cells can result in cell death or prevent the cell from passing through meiosis. Damage produced in postmeiotic cells, the spermatids or sperm are more likely to be transmitted to the F1 progeny (Albanese, 1987). In females it is during early fetal development of the ovary that oocyte stage is the most commonly tested stage in the adult female. To test other stages during the first or second meiotic divisions demands the use of oocytes undergoing ovulation which occur naturally or are hormone-stimulated. It is thus more difficult technically to test female germ cells.

Heritable Chromosome Assays Damage may be analyzed in the heritable translocation test, which involves the examination in male F1 animals if diakinesis metaphase 1 spermatocytes for multivalent association fall within the acceptable range for the laboratory for a substance to be considered positive or negative under the conditions of the study.

Germ Cell Cytogenetic Assays Either mouse or rat can be used but the mouse is generally the preferred species. Normally such assays are not conducted for routine screening purposes. Spermatogonial metaphases can be prepared by the air-drying technique of Evans and O’Riordan (1975) for the first and second meiotic metaphase (MI and MII) in the male mouse. This method is not so suitable for rat and hamster. The numbers of spermatogonial metaphases can be boosted if, prior to hypotonic treatment, the testicular tubules are dispersed in trypsin solution (0.25%). At least 1 month between treatment and sample should be allowed to pass in the mouse to allow treated cells to reach meiosis. Brook and Chandley (1986) established that 11 days and 4 hr was required for spermatogonial cells to reach preleptotene and 8 days and 10 hr to reach zygotene. It takes 4 hr for cells to move from MI to MII but test compounds can alter this rate. A search for multivalent formation can be made at MI for the structural rearrangements induced in spermatogonia. Cawood and Breckon (1983) examined the synaptonemal complex at pachytene, using electron microscopy. Errors of segregation should be searched for at the first meiotic division in the male mouse, MII cells showing 19 (hypoploid) and 21 (hyperploid) chromosomes (Brook and Chandley, 1986). Hansmann and El-Nahass (1979), Brook (1982), and Brook and Chandley (1985) describe assays in the female mouse and procedures used for inducing ovulation by hormones and treatment of specific stages of meiosis.

Sister Chromatid Exchange Assays Sister Chromatid Exchange assays (SCEs) are reciprocal exchanges between sister chromatids. They result in a change in morphology of the chromosome but breakage and reunion are involved although the exact mechanism is unclear. They are thought to occur at homologous loci. In 1958 Taylor demonstrated SCEs, using autoradiographic techniques to detect the disposition or labeled

Genotoxicity

175

DNA following incorporation of [3H]-thymidine. 5-Bromo2-deoxyuridine (drdU) has now replaced [3H]-thymidine and various staining methods have been used to show the differential incorporation of BrdU between sister chromatids: fluorescent Hoechst 33258 (Latt, 1973); combined fluorescent and Giemsa (Perry and Wolff, 1974); and Giemsa (Korenberg and Freedlander, 1974). The fluorescent plus Giemsa procedure is recommended in view of the fact that stained slides can be stored and microscope analysis is simpler. So that SCEs can be seen at metaphase, cells must pass through S phase (Kato, 1973, 1974; Wolff et al., 1974). SCEs appear to occur at the replication point, since SCE induction is maximal at the beginning of DNA synthesis but drops to zero at the end of S phase (Latt and Loveday, 1978). For SCE analysis in vitro, any cell type that is replicating or can be stimulated to divide is suitable. The incorporation of BrdU into cells in vivo allows the examination of a variety of tissues (Latt et al., 1980). Edwards et al. (1993) suggest that is necessary to standardize protocols measuring SCE since different responses can be obtained depending on the extent of simultaneous exposure of test compound and BrdU.

Relevance of SCE in Terms of Genotoxicity SCEs do not appear to be related to other cytogenetic events, since potent clastogens such as bleomycin and ionizing radiation induce low levels of SCE (Perry and Evans, 1975). The mechanisms involved in chromosome aberrations and SCE formation are dissimilar (e.g., Galloway and Wolff, 1979). There is no evidence that SCEs are in themselves lethal events, since there is little relationship to cytotoxicity (e.g., Bowden et al., 1979). It was suggested by Wolff (1977a, b) that they relate more to mutational events due to a compatibility with cell survival. However, there are examples of agents that induce significant SCE increases in the absence of mutation (Bradley et al., 1979) as well as they converse (Connell, 1979; Connell and Medcalf, 1982). The SCE assay is particularly sensitive for alkylating agents and base analogues, agents causing single-strand breaks in DNA and compounds acting through DNA binding (Latt et al., 1981). The most potent SCE inducers are S-phase-dependent. Painter (1980) reports that agents such as X-irradiation, which inhibits replicon initiation, are poor SCE inducers, whereas mitomycin C, which inhibits replication fork progression, is a potent SCE inducer.

Experimental Design Established cell lines, primary cell cultures of rodents may be used. Detailed information on in vitro and in vivo assays may be obtained in reviews of SCE methods by Latt et al. (1977, 1981), Perry and Thompson (1984) and Perry et al. (1984). The in vitro methods will be briefly explored here. Either monolayer or suspension cultures can be employed, or human lymphocytes. Human fibroblasts are less suitable because of their long cell cycle duration. The concentration of organic solvents for the test compound should not exceed 0.8% v/v, as higher concentrations could lead to slight elevations in the SCE level (Perry et al., 1984). For monolayer cultures, the cultures are set up the day before BrdU treatment so that the cells will be in exponential growth before the addition of BrdU or the test compound. After BrdU addition the cells are allowed to undergo the

176

Safety Evaluation of Medical Devices

equivalent of two cell cycles before cell harvest. A spindle inhibitor such as colchicine or colcemid is introduced for the final 1–2 hr of culture to arrest cells in metaphase, after which the cells are harvested and chromosome preparations are made by routine cytogenetic techniques. In the absence of metabolic activation, BrdU and the test agent can be added simultaneously and left for the duration of BrdU labeling. Shorter treatments should be used in the presence of metabolic activation or to avoid synergistic effects with BrdU, when cells can be pulse treated for, e.g., 1 hr before BrdU addition (Edwards et al., 1993). Peripheral blood cultures are established in medium containing BrdU and PHA. Cocemid is added 1–2 hr before harvest and the cells are harvested between 60 and 70 hr post PHA stimulation. Cell harvest and slide preparations are conducted according to routine cytogenetic methods. Heparinized blood samples may be stored at 4ºC for up to 48 hr without affecting the SCE response (Lambert et al., 1982). If the test agent is known to react with serum or red blood cells, the mononuclear lymphocytes may be isolated by use of a Ficoll/Hypaque gradient (Boyum, 1968). If metabolic activation is not required, treatment is best conducted over the whole of the final 24 hr of culture, or if metabolic activation is required, a pulse exposure may be employed to treat cultures at the first S phase at around 24–30 hr, or at 48 hr for an asynchronous population. Exposure of cells to fluorescent light during the culture period leads to photolysis of BrdU-containing DNA and a concomitant increase in SCE frequency (Wolff and Perry, 1974). Consequently, SCE cultures should be kept in the dark and manipulated under subdued light conditions such as yellow safe light. Furthermore, media used in SCE assays should be stored in the dark, since certain media components produce reactive SCE-inducing intermediates on exposure to fluorescent light (Monticone and Schneider, 1979). Coded and randomized slides should be read. All experiments should be repeated at least once (Perry et al., 1984) with higher and lower concentrations of S9 mix if a negative response is achieved. Even for an apparently unambiguous positive response with a greater than twofold increase in SCEs over the background level at the highest dose, and with at least two consecutive dose levels with an increased SCE response, a repeat study is necessary to show a consistent response. The quality of differential staining will determine the ease and accuracy of SCE scoring, and, to eliminate variation, results from different observers should occasionally be compared. Furthermore, to avoid observer bias, scorers should have slides from different treatment groups equally distributed among them, as with all cytogenetic studies.

REFERENCES Adler, I.D. and Ashby, J. (1989). The present lack of evidence for unique rodent germ-cell mutagens. Mutation Res., 212: 55–66. Advisory Committee on Dangerous Pathogens (1984). Categorisation of Pathogens according to Hazard and Categories of Containment. London: HMSO. Albanese, R. (1987). Mammalian male germ cell cytogenetics. Mutagenesis, 2: 79–85. Amacher, D.E., Paillet, S.C., Turner, G.N., Ray, V.A. and Salsburg, D.S. (1980). Point mutations at the thymidine kinase locus in L5178Y

mouse lymphoma cells. 2. Test validation and interpretation. Mutation Res., 72: 447–474. Ames, B.N. (1971). The detection of chemical mutagens with enteric bacteria. In: A. Hollaender, (Ed.), Chemical Mutagens, Principles and Methods for Their Detection, Vol. 1. Plenum Press, New York., pp. 267–282. Ames, B.N., McCann, J. and Yamasaki, E. (1975), Methods for detecting carcinogens and mutagens with the Salmonella/ mammalian-microsome mutagenicity test. Mutat. Res. 31: 237–364. Anderson, D. (1990). Male mediated F1 abnormalities. Mutation Res., 229: 103–246. Applegate, M.L., Moore, M.M., Broder, C.B. et al. (1990). Molecular dissection of mutations at the heterozygous thymidine canise locus in mouse lymphoma cells. Proc. Natl. Acad. Sci. USA, 87: 51–55. Arlett, C.F. and Cole, J. (1990). The third United Kingdom Environmental Mutagen Society collaborative trial: overview, a summary and assessment. Mutagenesis, 5 (Suppl.): 85–88. Arlett, C.F., Smith, D.M., Clark, G.M., Green, J.H.L., Cole, J., McGregor, D.B. and Asquith, J.C. (1989). Mammalian cell assays based upon colony formation. In: D.J. Kirkland, (Ed.), UKEMS Subcommittee on Guidelines for Mutagenicity Testing. Report Part III: Statistical Evaluation of Mutagenicity Test Data. Cambridge University Press, Cambridge, pp. 66–101. Ashby, J. and Tennant, R.W. (1988). Chemical structure, Salmonella mutagenicity and extent of carcinogenicity as indices of genotoxic carcinogens among 222 chemicals tested in rodents by the US NCI/NTP. Mutation Res., 204: 17–115. Ashby, J., Tennant, R.W., Zeiger, E. and Stasiewiczs, S. (1989). Classification according to chemical structure, mutagenicity to Salmonella and level of carcinogenicity of a further 42 chemicals tested for carcinogenicity by the U.S. National Toxicology Program. Mutation Res., 223: 73–104. Ashwood-Smith, M.J. (1980). Stability of frozen microsome preparations for use in the Ames Salmonella mutagenicity assay. Mutation Res., 69: 199–200. Bartsch, H., Camus, A.-M. and Malaveille, C. (1976). Comparative mutagenicity of N-nitrosamines in a semi-solid and in a liquid incubation system in the presence of rat or human tissue fractions. Mutation Res., 37: 149–162. Beckman, D.A. and Brent, R.L. (1986). Mechanism of known environmental teratogens: drugs and chemicals. Clin. Perinatol., 13: 649–687. Bender, M.A., Griggs, H.G. and Bedford, J.S. (1974). Mechanisms of chromosomal aberration production. III. Chemicals and ionizing radiation. Mutation Res., 23: 197–212. Blazak, W.F., Steward, B.E., Galperin, I., Allen, K.L., Rudd, C.J., Mitchell, A.D. and Caspary, W.J. (1986). Chromosome analysis of triflourothymidine-resistant L5178Y mouse lymphoma cells colonies. Environ. Mutagen., 8: 229–240. Bossi, L. (1985). Information suppression. In: J., Scaife, D. Leach, and A. Galizzi, (Eds), Genetics of Bacteria. Academic Press, New York., pp. 49–64. Bowden, G.T., Hsu, I.C. and Harris, C.C. (1979). The effect of caffeine on cytotoxicity, mutagenesis and sister chromatid exchanges in Chinese hamster cells treated with dihydrodiol epoxide derivatives of benzo(a)pyrene. Mutation Res., 63: 361–370. Boyum, A. (1968). Separation of lymphocytes and erythrocytes by centrifugation, Scand. J. Clin. Invest., 21: 77–85. Bradley, M.O., Hsu, I.C. and Harris, C.C. (1979). Relationships between sister chromatid exchange and mutagenicity, toxicity and DNA damage, Nature, 282: 318–320. Braithwaite, I. and Ashby, J. (1988). A non-invasive micronucleus assay in rat liver. Mutation Res., 203: 23–32. Bridges, B.A. (1972). Simple bacterial systems for detecting mutagenic agents. Lab. Pract., 21: 413–419. Bridges, B.A. (1976). Short-term screening tests for carcinogens. Nature, 261: 195–200. Bridges, B.A., Woodgate, R., Ruiz-Rubio, M., Sharif, F., Sedgwick, S.G. and Hubschere, U. (1987). Current understanding of UV-induced base pair substitution mutation in E. coli with particular reference to the DNA polymerase III complex. Mutation Res., 181: 219–226.

Chapter 12: Brook, J.D. (1982). The effect of 4CMB on germ cells of the mouse, Mutation Res., 100: 305–308. Brook, J.D. and Chandley, A.C. (1985). Testing of 3 chemical compounds for aneuploidy induction in the female mouse, Mutation Res., 157: 215–220. Brook, J.D. and Chandley, A.C. (1986). Testing for the chemical induction of aneuploidy in the male mouse. Mutation Res., 164: 117–125. Brusick, D. (1986). Genotoxic effects in cultures mammalian cells produced by low pH treatment conditions and increased ion concentrations. Environ. Mutagen., 8: 879–886. Brusick, D. (1987a). Genotoxicity produced in cultures mammalian cell assays by treatment conditions. Special issue. Mutation Res., 189: 1–80. Brusick, D. (1987b). Principles of Genetic Toxicology, 2nd edn. Plenum Press, New York., London. Burnett, C., Fuchs, C., Corbett, J. and Menkart, J. (1982). The effect of dimethylsulphoxide on the mutagenicity of the hair-dye, p-phenylenediamine. Mutation Res., 103: 1–4. Butterworth, B.E. and Slaga, T.J. (1987). Nongenotoxic Mechanisms in Carcinogenesis. Banbury Report No. 25. Cold Spring Harbor Laboratory, New York. Carter, T.C., Lyon, M.F. and Philips, R.J.S. (1956). Induction of mutations in mice by chronic gamma irradiation; interim report. Br. J. Radiol., 29: 106–108. Carver, J.H., Machado, M.L. and MacGregor, J.A. (1985). Petroleum distillates suppress in vitro metabolic activation: Higher (S9). Required in the Salmonella/microsome mutagenicity assay. Environ. Mutagen., 7: 369–380. Cawood, A.D. and Breckon, G. (1983). Synaptonemal complexes as indicators of induced structural change in chromosomes after irradiation of spermatogonia. Mutation Res., 122: 149–154. Chandley, A.C. (1981). The origin of chromosomal aberrations in man and their potential for survival and reproduction in the adult human population. Ann. Genet., 24: 5–11. Chu, E.H.Y. and Malling, H.U. (1968). Mammalian cell genetics. II. Chemical induction of specific lucus mutations in Chinese hamster cells in vitro. Proc. Natl Acad. Sci. USA, 61: 1306–1312. Cifone, M.A., Myhr, B., Eiche, A. and Bolisfodi, G. (1987). Effect of pH shifts on the mutant frequency at the thymidine kinase locus in mouse lymphoma L5178Y TK+/− cells, Mutation Res., 189: 39–46. Clayson, D.B. (1989). ICPEMC publication No. 17: Can a mechanistic rationale be provided for non-genotoxic carcinogens identified in rodent bioassays? Mutation Res., 221: 53–67. Cleaver, J.E. (1983). Xeroderma pigmentosum. In: J.B., Stanbury, J.B., Wyngaarden, D.S., Fredrickson, J.C. Goldstein, and M.S. Brown, (Eds), The Metabolic Basis of Inherited Disease. McGraw-Hill, New York., pp. 1227–1248. Clive, D. (1987). Historical overview of the mouse lymphoma TK+/− mutagenicity assay. In M.M., Moore, D.M., Demarini, F.J. De Serres, and K.R. Tindall, (Eds), Mammalian Cell Mutagenesis. Banbury Report 28. Cold Spring Harbor Laboratory, New York, pp. 25–36. Clive, D. (1989). Mammalian cell genotoxicity: A major role for non-DNA targets? Mutation Res., 223: 327–328. Clive, D., Caspary, W., Kirkby, P.E., Krehl, R., Moore, M., Mayo, J. and Oberly, T. J. (1987). Guide for performing the mouse lymphoma assay for mammalian cell mutagenicity. Mutation Res., 189: 145–146. Clive, D., Flamm, W. G. and Patterson, J. B. (1972). A mutational assay system using the thymidine kinase locus in mouse lymphoma cells. Mutation Res., 16: 77–87. Cole, J., Fox, M., Garner, R.C., McGregor, D.B. and Thacker, J. (1990). Gene mutation assays in cultured mammalian cells. In: D.J. Kirkland, (Ed.), UKEMS Subcommittee on Guidelines for Mutagenicity Testing. Report Part I revised. Cambridge University Press, Cambridge, pp. 87–114. Cole, J., Muriel, W.J. and Bridges, B.A. (1986). The mutagenicity of sodium flouride to L5178Y (wildtype and TK+/− 3.7.2c) mouse lymphoma cells. Mutagenesis, 1: 157–167. Connell, J.R. (1979). The relationship between sister chromatid exchange, chromosome aberration and gene mutation induction

Genotoxicity

177

by several reactive polycyclic hydrocarbon metabolites in cultured mammalian cells, Int. J. Cancer, 24: 485–489. Connell, J.R. and Medcalf, A.S. (1982). The induction of SCE and chromosomal aberrations with relation to specific base methylation of DNA in Chinese hamster cells by N-methyl-n-nitrosourea and dimethyl sulphate, Carcinogenesis, 3: 385–390. Cullum, J. (1985) Insertion sequences, In: Scaife, J., Leach, D., and Galizzi, A. (Eds), Genetics of Bacteria, Academic Press. New York, pp. 85–96. Dean, B.J. and Danford, N. (1984). Assays for the detection of chemically-induced chromosome damage in cultures mammalian cells. In: S. Venitt, and J.M. Parry, (Eds), Mutagenicity Testing. A Practical Approach. IRL Press, Oxford, pp. 187–232. DeSerres, F.J. and Shelby, M.D. (1979). Recommendations on datas production and analysis using the Salmonella/microsome mutagenicity assay. Mutation Res., 64: 159–165. deVries, J.H. (1901). The Mutation Theory, Verlag von Veit & Co., Leipzig. Doll, R. (1977). Strategy for detection of cancer hazards to man. Nature, 265: 589–596. Dorado, G., Ariza, R.R. and Pueyo, C. (1988). Mutagenicity of red wine in the L-arabinose resistance test with Salmonella typhimurium. Mutagenesis, 3: 497–502. Dunnett, C.W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc., 50: 1096–1121. Edwards, A.J., Moon, E.Y., Anderson, D. and McGregor, D.B. (1993). The effect of simultaneous exposure to bromodeoxyuridine and methyl methansulphonate on sister chromatid exchange frequency in culture human lymphocytes and its mutation research. Mutation Res., 247: 117–125. Ehling, U.H., Chu, E.H.Y., DeCarli, L., Evans, H.J., Hayashi, M., Lambert, B., Neubert, D., Thilly, W.G. and Vainio, H. (1986). Report 8. Assays for germ-cell mutations in mammals. In: R. Montesano, H. Bartsch, H. Vainio, J. Wilbourn, and H. Yamasaki, (Eds), Long-term and Short-term Assays for Carcinogens. A Critical Appraisal. IARC Scientific Publications, No. 83, Lyon, pp. 245–265. Evans, H.H., Mencl, J., Horng, M.F., Ricanti, M., Sanchez, D. and Hozier, J. (1986). Lucus specificity in the mutability of L5178Y mouse lymphoma cells: the role of multilocus lesions. Proc. Natl Acad. Sci. USA, 83: 4379–4385. Evans, H.J. and O’Riordan, M.L. (1975). Human peripheral blood lymphocytes for the analysis of chromosome aberrations in mutagen tests. Mutation Res., 31: 135–148. Ford, D.K. and Yerganian, G. (1958). Observations on the chromosomes of Chinese hamster cells in tissue culture. J. Natl Cancer Inst., 21: 393–425. Forster, R., Green, M.H.L. and Priestley, A. (1980). Optimal Levels of S9 fraction in Ames and fluctuation tests: apparent importance of diffusion of metabolites from top agar. Carcinogenesis, 2: 1081–1085. Fox, M. (1981). Some quantitative aspects of the response of mammalian in vitro to induced mutagenesis. In: J.J. Marchelonis and M.G. Hanna (Eds.), Cancer Biology Reviews, Vol. 3. Marcel Dekker, New York, pp. 23–62. Friedberg, E.C. (1988). DNA repair in the yeast Saccharomyces cerevisiae. Microb. Rev., 52: 70–102. Galloway, S.M. and Wolff, S. (1979). The relation between chemically induced SCE’s and chromatid breakage. Mutation Res., 61: 297–307. Gatehouse, D.G. and Delow, G.F. (1979). The development of a ‘MicrotitreR’ fluctuation test for the detection of indirect mutagens and its use in the evaluation of mixed enzyme induction of the liver. Mutation Res., 60: 239–252. Gatehouse, D.G. and Tweats, D.J. (1987). Letter to the Editor. Mutagenesis, 1: 307–308. Gatehouse, D. and Wedd, D.J. (1984). The differential mutagenicity of isoniazid in fluctuation assays and Salmonella plate tests. Carcinogenesis, 5: 391–397. Gatehouse, D.G., Wedd, D.J., Paes, D., Delow, G., Burlinson, B., Pascoe, S., Brice, A., Stemp, G. and Tweats, D.J. (1988). Investigations into the genotoxic potential of Ioxtidine, a long-acting H2-receptor antagonist. Mutagenesis, 3: 57–68.

178

Safety Evaluation of Medical Devices

Gatehouse, D.G., Wilcox, P., Forster, R., Rowland, I.R. and Callander, R.D. (1990). Bacterial mutation assays. In: D.J. Kirkland, (Ed.), Basic Mutagenicity Tests: UKEMS Recommended Procedures. Cambridge University Press, Cambridge, pp. 13–61. Green, M.H.L. and Muriel, W.J. (1976). Mutagen testing using TRP+ reversion in E. coli. Mutation Res., 38: 3–32. Hansmann, I. and El-Nahass, E. (1979). Incidence of non-disjunction in mouse oocytes. Cytogenet. Cell Genet., 24: 115–121. Haworth, S., Lawlor, T., Mortelmanns, K., Speck, W. and Zeiger, E., (1983). Salmonella mutagenicity results for 250 chemicals. Environ. Mutagen. Suppl., 1: 3–142. Heddle, J.A. and Bodycote, D.J. (1970). On the formation of chromosomal aberrations. Mutation Res., 9: 117–126. Hera, C. and Pueyo, C. (1986). Conditions for optimal use of the L-arabinose-resistance mutagenesis test with Salmonella typhimurium. Mutagenesis, 1: 267–274. Herrera, G., Urios, A., Alexandre, V. and Blanco, M. (1988). UV light induced mutability in Salmonella strains containing the umu DC or the muc AB operon: evidence for a umu C function. Mutation Res., 198: 9–13. Holden, H.E. (1982). Comparison of somatic and germ cell models for cytogenetic screening. J. Appl. Toxicol., 2: 196–200. Hozier, J., Sawger, D., Clieve, D. and Moore, M. (1982). Cytogenetic distinction between the TK+ and TK− chromosomes in L5178Y/TK+/− -3.7.2.C mouse lymphoma cell line. Mutation Res., 105: 451–456. Hsie, A.W., Brimer, P.A., Mitchell, T.J. and Gosslee, D.G. (1975). The dos-response relationship for ethyl methane sulfonate-induced mutation at the hypoxanthine-guanine phosphoribosyl transferase locus in Chinese hamster ovary cells. Somatic Cell Genet., 1: 247–261. Hubbard, S.A., Brooks, T.M., Gonzalez, L.P. and Bridges, J.W. (1985). Preparation and characterization of S9-fractions. In: J.M. Parry, and C.F. Arlett, (Eds), Comparative Genetic Toxicology. Macmillan, London, pp. 413–438. IARC (1979). Handling Chemical Carcinogens in the Laboratory; Problems of Safety, Scientific Publications, No. 33. International Agency for Research on Cancer, Lyons, France. ICEMC (1983). Committee and Final Report: screening strategy for chemicals that are potential germ-cell mutagens in mammals, Mutation Res., 114, 117–177. ISCN (1985). An International System for Human Cytogenetic Nomenclature. In: D.G. Harnden and H.P. Klinger (Eds.), Report of the Standing Committee on Human Cytogenetic Nomenclature. Karger, Basel, Switzerland. Ishidate, M. Jr. (1988a). Data Book of Chromosomal Aberration Tests in vitro. Elsevier, Amsterdam, The Netherlands. Ishidate, M. Jr. (1988b). A proposed battery of tests for the initial evaluation of the mutagenic potential of medicinal and industrial chemicals. Mutation Res., 205: 397–407. Ishidate, M. Jr. and Harnois, M.C. (1987). The clastogenicity of chemicals in mammalian cells. Letter to the editor, Mutagenesis, 2: 240–243. Ishidate, M. Jr., Sofuni, T., Yoshikawa, K. et al. (1984). Primary mutagenicity screening of food additives currently used in Japan. Fd. Chem. Toxicol., 22: 623–636. ISO (1993). Biological Evaluation of Medical Devices—Part 3: Tests for Genotoxicity, Carcinogenicity, and Reproductive Toxicity. ISO 10993-3. JMHW (1984). Guidelines for Testing of Drugs for Toxicity. Pharmaceutical Affairs Bureau, Notice No. 118. Ministry of Health and Welfare, Japan. Joint Directives of the Japanese Environmental Protection Agency, Japanese Ministry of Health and Welfare and Japanese Ministry of International Trade and Industry, 31 March 1987. Jordan, E., Saedler, H. and Starlinger, P. (1967). Strong polar mutations in the transferase gene of the galactose operon in E. coli. Molec. Gen. Genet., 100: 296–306. Kada, T. (1981). The DNA damaging activity of 42 coded compounds in the Rec-assay. In: F. de Serres and J. Ashby (Eds.), Evaluation of Short-term Tests for Carcinogens—Report of the International Collaborative Program. Elsevier/North Holland, Amsterdam, The Netherlands, pp. 175–182.

Kakunaga, T. and Yamasaki, H. (Eds) (1984). Transformation Assays of Established Cell Lines: Mechanisms and Application. Proceedings of a Workshop Organized by IARC in Collaboration with the U.S. National Cancer Institute and the U.S. Environmental Protection Agency, Lyon 15–17 Feb. 1984. IARC Scientific Publication No. 67. Kalter, K. (1977). Correlation between teratogenic and mutagenic effects of chemicals in mammals. In: A. Hollanender (Ed.), Chemical Mutagens: Principles and Methods for Their Detection, Vol. 6. Plenum Press, New York, pp. 57–82. Kato, H. (1973). Induction of sister chromatid exchanges by UV light and its inhibition by caffeine. Exptl. Cell Res., 82: 383–390. Kato, H. (1974). Induction of sister chromatid exchanges by chemical mutagens and its possible relevance to DNA repair. Exptl. Cell Res., 85: 239–247. Kato, T. and Shinoura, Y. (1977). Isolation and characterization of mutants of Escherichia coli deficient in induction of mutation by ultraviolet light. Molec. Gen. Genet., 156: 121–132. Kennelly, J.C., Stanton, C. and Martin, C.N. (1984). The effect of acetyl-CoA supplementation on the mutagenicity of benzidines in the Ames assay. Mutation Res., 137: 39–45. Kirkland, D.J. and Garner, R.C. (1987). Testing for genotoxicitychromosomal aberrations in vitro—CHO cells or human lymphocytes? Mutation Res., 189: 186–187. Korenberg, J.R. and Freedlender, E.F. (1974). Giesma technique for the detection of sister chromatid exchanges, Chromasoma, 48: 355–360. Kuroki, T. and Matsushima, T. (1987). Performance of short-term tests for detection of human carcinogens. Mutagenesis, 2(1): 33–37. Lambert, B., Lindblad, A., Holmberg, K. and Francesconi, D. (1982). The use of sister chromatid exchange to monitor human populations for exposure to toxicologically harmful agents. In: S. Wolff (Ed.), Sister Chromatid Exchange Wiley. New York, pp. 149–182. Latt, S.A. (1973). Microfluorometric detection of deoxyribonucleic acid replication in human metaphase chromosomes. Proc. Natl. Acad. Sci. USA, 770: 3395–3399. Latt, S., Allen, J.W., Bloom, S.E. et al. (1981). Sister chromatid exchanges: a report of the gene-tox program. Mutation Res., 87: 17–62. Latt, S.A., Allen, J.W., Rogers, W.E. and Jurgens, L.A. (1977). In vitro and in vivo analysis of sister chromatid exchange formation. In: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.), Handbook of Mutagenicity Test Procedures. Elsevier, Amsterdam, The Netherlands, pp. 275–291. Latt, S.A. and Loveday, K.S. (1978). Characterization of sister chromatid exchange induction by 8-methoxypsoralen plus UV light. Cytogenet. Cell Genet., 21: 184–200. Latt, S.A., Schreck, R.R., Loveday, K.S., Dougherty, C.P. and Shuler, C.F. (1980). Sister chromatid exchanges, Adv. Human Genet., 31: 291–298. Lawley, P. (1989). Mutagens as carcinogens: development of current concepts. Mutation Res., 213: 3–26. Li, A.P., Aaron, C.S., Aueltta, A.E., Dearfield, K.L., Riddle, J.C., Slesinski, R.S. and Stankowski, L.F. Jr. (1991). An evaluation of the roles of mammalian cell mutation assays in the testing of chemical genotoxicity. Regulatory Toxicol. Pharmacol., 14: 24–40. Li, A.P., Carver, J.H., Choy, W.N., Bupta, R.S., Loveday, K.S., O’Neill, J.P., Riddle, J.C., Stankowski, L.F. and Yang, L.C. (1987). A guide for the performance of the Chinese Hamster ovary cell/hypoxanthine guanine phosphoribosyl transferase gene mutation assay. Mutation Res., 189: 135–141. Linblad, W.J. and Jackim, E. (1982). Mechanism for the differential induction of mutation by kS9 activated benzo(a)pyrene employing either a glucose-6-phosphate dependent NADPHregenerating system or an isocitrate dependent system. Mutation Res., 96: 109–118. Lindahl, T. (1979). DNA glycoslylases, endonucleases for apurinic/ apyrimidinic sites and base excision repair. Proc. Nucl. Acid Res. Mol. Biol., 22: 109–118. Lindahl, T., Sedwick, B., Sekiguchi, M. and Nakabeppu, Y. (1988). Regulation and expression of the adaptive response to alkylating agents. Ann. Rev. Biochem., 57: 133–157.

Chapter 12: Little, J.W. (1984). Autodigestion of lex A and phage T repressors, Proc. Natl. Acad. Sci. USA, 81: 1375–1379. Little, J.W. and Mount, D.W. (1982). The SOS regulatory system of Escherichia coli. Cell, 29: 11–22. MacGregor, J.T., Wehr, C.M. and Gould, D.H. (1980). Clastogeninduced micronuclei in peripheral blood erythrocytes: the basis of an improved micronucleus test. Environ. Mutagen., 2: 509–514. MacPhee, D.G. (1973). Salmonella typhimurium hisG46 (R-Utrecht): possible use in screening mutagens and carcinogens. Appl. Microbiol., 26: 1004–1005. Madle, S. and Obe, G. (1980). Methods for the analysis of the mutagenicity of indirect mutagens/carcinogens in eukaryotic cells. Human Genet., 56: 7–20. Mahon, G.A.T., Green, M.H.L., Middleton, B., Mitchell, I. DeG., Robinson, W.D. and Tweats, D.J. (1989). Analysis of data from microbial colon assays. In: D.J. Kirkland (Ed.), Statistical Evaluation of Mutagenicity Test Data. Cambridge University Press, Cambridge, England, pp. 26–65. Marinus, M.G. and Morris, R.N. (1974). Biological function for the 6-methyladenine residues in the DNA of Escherichia coli K12. J. Molec. Biol., 85: 309–322. Maron, D.M. and Ames, B.N. (1983). Revised methods for the Salmonella mutagenicity test. Mutation Res., 113: 173–215. Miller, E.C. and Miller, J.A. (1971). The mutagenicity of chemical carcinogens: correlations, problems and interpretations. In: A. Hollaender, (Ed.), Chemical Mutagens, Principles and Methods for Their Detection, Vol. 1. Plenum Press, New York., pp. 83–120. Miller, J.H. (1985). Pathways of mutagenesis revealed by analysis of mutational specificity. In: J. Scaife, D. Leach and A. Galizzi (Eds.), Genetics of Bacteria. Academic Press, New York, pp. 25–40. Modrich, P. (1987). DNA mismatch correction. Ann. Rev. Biochem., 56: 435–466. Monticone, R.E. and Schneider, E.L. (1979). Induction of SCEs in human cells by fluorescent light. Mutation Res., 59: 215–221. Morimoto, K., Sato, M. and Koizumi, A. (1983). Proliferative kinetics of human lymphocytes in culture measure by autoradiography and sister chromatid differential staining. Epithel. Cell Res., 145: 249–356. Mortelmanns, K.E. and Dousman, L. (1986). Mutagenesis and plasmids. In: F.J. de Serres (Ed.), Chemical Mutagens, Principles and Methods for Their Detection, Vol. 10. Plenum Press, New York, pp. 469–508. Mortelmanns, K.E. and Strocker, B.A.D. (1979). Segregation of the mutator property of plasmid R46 from its ultraviolet-protecting property. Molec. Gen. Genet., 167: 317–328. MRC (1981). Guidelines for Work with Chemical Carcinogens in Medical Research Council Establishments. Medical Research Council, London. Muller, H.J. (1927). Artificial transmutation of the gene. Science, 66: 84–87. Natarajan, A.T. and Obe, G. 1982). Mutagenicity testing with cultured mammalian cells: cytogenetic assays. In: J.A. Heddle, (Ed.), Mutagenicity, New Horizons in Genetic Toxicology. Academic Press, New York., pp. 172–213. Neudecker, T. and Henschler, D. (1985). Allyl isothiocyanate is mutagenic in Salmonella typhimurium. Mutation Res., 30: 143–148. Newton, M.F. and Lilly, L.J. (1986). Tissue specific clastogenic effects of chromium and selenium salts in vivo. Mutation Res., 169: 61–69. Oberly, T.J., Bewsey, B.J. and Probst, G.S. (1987). A procedure for the CHO/HGPRT mutation assay involving treatment of cells in suspension culture and selection of mutants in soft agar. Mutation Res., 182: 99–111. OECD (1983). OECD Guidelines for the Testing of Chemicals. No. 475. Genetic toxicology: in vivo mammalian bone marrow cytogenetic test—chromosomal analysis. Adopted 4 April 1984. Ong. T., Mukhtar, M., Wolf, C.R. and Zeiger, E. (1980). Differential effects of cytochrome P450-inducers on promutagen activation capabilities and enzymatic activities of S-9 from rat liver. J. Environ. Pathol. Toxicol., 4: 55–65. Pagano, D.A. and Zeiger, E. (1985). The stability of mutagenic chemicals tested in solution. Environ. Mutagen., 7: 293–302.

Genotoxicity

179

Painter, R.B. (1980). A replication model of sister-chromatid exchange. Mutation Res., 70: 337–341. Parry, J.M., Arlett, C.F. and Ashby, J. (1985). An overview of the results of the in vivo and in vitro test systems used to assay the genotoxicity of BZD, DAT, DAB and CDA in the second UKEMS study. In: J.M. Parry, and C.F. Arlett, (Eds), Comparative Genetic Toxicology: The Second UKEMS Collaborative Study. Macmillan, London, pp. 597–616. Perry, P.E. and Evans, H.J. (1975). Cytotological detection of mutagen/carcinogen exposure by sister chromatid exchange. Nature, 258: 121–125. Perry, P., Henderson, L. and Kirkland, D. (1984). Sister chromatid exchange in cultured cells. In: UKEMS Subcommittee on Guidelines for Mutagenicity Testing. Report Part IIA, pp. 89–121. Perry, P.E. and Thomson, E.J. (1984). Sister chromatid exchange methodology. In: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.), Handbook of Mutagenicity Test Procedures. Elsevier, Amsterdam, Netherlands, pp. 495–529. Perry, P.E. and Wolff, S. (1974). New Giemsa method for the differential staining of sister chromatids. Nature, 251: 156–158. Pinney, R.J. (1980). Distribution among incompatibility groups of plasmids that confer UV mutability and UV resistance. Mutation Res., 72: 155–159. Preston. R.J., Au, W., Bender, M. et al. (1981). Mammalian in vivo and in vitro cytogenetic assays. Mutation Res., 87: 143–188. Preston, R.J., San Sebastian, J.R. and McFee, A.F. (1987). The in vitro human lymphocyte assay for assessing the clastogenicity of chemical agents. Mutation Res., 189: 175–183. Prival, M.J., Bell, S.J., Mitchell, V.D., Peiperi, M.D. and Vaughn, V.L. (1984). Mutagenicity of benzidine and benzidine-congener dyes and selected monoazo dyes in a modified Salmonella assay. Mutation Res., 136: 33–47. Prival, M.J. and Mitchell, V.D. (1982). Analysis of a method for testing azo-dyes for mutagenic activity in S. typhimurium in the presence of FMN in hamster liver S9. Mutation Res., 97: 103–116. Pueyo, C. and Lopez-Barea, J. (1979). The L-arabinose-resistance test with Salmonella typimurium strain SV3 selects forward mutations at several ara genes. Mutation Res., 64: 249–258. Revell, S.H. (1974). The breakage-and–reunion theory and the exchange theory for chromosome aberrations induced by ionizing radiations: A short history. In: J.T. Lett and M. Zelle. (Eds.), Advances in Radiation Biology, Vol. 4. Academic Press, New York, pp. 367–415. Richardson, C., Williams, D.A., Allen, J.A., Amphlett, G., Changer, D.O. and Phillips, B. (1989). Analysis of data from in vitro cytogenetic assays. In: D.J. Kirkland, (Ed.), UKEMS Sub-committee on Guidelines for Mutagenicity Testing. Report. Part III. Statistical Evaluation of Mutagenicity Test Data. Cambridge University Press, Cambridge, pp. 141–154. Robinson, W.D., Green, M.H.L., Cole, J., Healy, M.J.R., Garner, R.C. and Gatehouse, D. (1989). Statistical Evaluation of bacterial mammalian fluctuation tests. In: D.J. Kirkland, (Ed.), Statistical Evaluation of Mutagenicity Test Data. Cambridge University Press, Cambridge, pp. 102–140. Rossman, T.G. and Klein, C.B. (1988). From DNA damage to mutation in mammalian cells: A review. Environ. Molec. Mutagen., 11: 119–133. Roth, J.R. (1974). Frameshift mutations. Ann. Rev. Genet., 8: 319–346. Russell, L.B. (1984). Procedures and evaluation of results of the mouse spot test. In: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.), Handbook of Mutagenicity Test Procedures. Elsevier, Amsterdam, The Netherlands, pp. 393–403. Russell, W.L. (1951). X-ray induced mutations in mice. Cold Spring Harbor Symp. Quant. Biol., 16: 327–336. Russell, W.L. (1989). Reminiscences of a mouse specific-locus addict. Environ. Molec. Mutagen., 14 (Suppl. 16): 16–22. Samson, L. and Cairns, J. (1977). A new pathway for DNA repair in E. coli. Nature, 267: 281–282. Scott, D., Danford, N., Dean, B., Kirkland, D. and Richardson, C. (1983). In vitro chromosome aberration assays. In: B.J. Dean (Ed.), UKEMS Subcommittee on Guidelines for Mutagenicity Testing, Report Part I, Basic Test Battery. UKEMS, Swansea, U.K., pp. 63–86.

180

Safety Evaluation of Medical Devices

Scott, D., Dean, B.J., Danford, N.D. and Kirkland, D.J. (1990). Metaphase chromosome aberration assays in vitro. In: D.J. Kirkland (Ed.), UKEMS Subcommittee on Guidelines for Mutagenicity Testing, Report. Part I: Revised Basic Mutagenicity Tests, UKEMS Recommended Procedures. Cambridge University Press, Cambridge, pp. 63–86. Searle, A.G. (1984). The specific locus test in the mouse. In: B.J. Kilbey, M. Legator, W. Nichols and C. Ramel (Eds.), Handbook of Mutagenicity Test Procedures. Elsevier, Amsterdam, The Netherlands, pp. 373–391. Sedgwick, S.G. and Goodwin, P.A. (1985). Differences in mutagenic and recombinational DNA repair in enterobacteria. Proc. Natl. Acad. Sci. USA, 82: 4172–4176. Selby, P.B. and Olson, W.H. (1981). Methods and criteria for deciding whether specific-locus mutation-rate data in mice indicates a positive, negative or inconclusive result. Mutation Res., 83: 403–418. Shanabruch, W.G. and Walker, G.C. (1980). Localization of the plasmid (pKM101) gene(s) involved in recA+lexA+-dependent mutagenesis. Molec. Gen. Genet., 129: 289–297. Steel, R.G.D. and Torrie, J.H. (1960). Principles and Procedures of Statistics. McGraw-Hill, New York. Streisinger, G., Okada, T., Emrich, J. et al. (1966). Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol., 31: 77–84. Tates, A.D., Neuteboom, I., Hofker, M. and den Engelese, L. (1980). A micronucleus technique for detecting clastogenic effects of mutagens/carcinogens (DEN, DMN) in hepatocytes of rat liver in vivo. Mutation Res., 74: 11–20. Taylor, J.H. (1958). Sister chromatid exchanges in tritium labeled chromosomes. Genetics, 43: 515–529. Tennant, R. W., Margolin, B. H., Shelby, M. D., Zeiger, E., Haseman, J. K., Spaulding, J., Caspary, W., Resenick, M., Statiewicz, S., Anderson, B. and Minor, R. (1987) Prediction of chemical carcinogenicity in rodents in vitro genetic toxicity assays, Science, 236: 933–941. The Collaborative Study Group for the Micronucleus Test (1986). Sex Differences in the micronucleus test. Mutation Res., 172: 151–163. The Collaborative Study Group for the Micronucleus Test (1988). Strain differences in the micronucleus test. Mutation Res., 204: 307–316. Thompson, E.D. (1986). Comparison of in vivo and in vitro cytogenetic assay results. Environ. Mutagen., 8: 753–767. Thust, R., Mendel, J., Schwarz, H. and Warzoki, R. (1980). Nitrosated urea pesticide metabolites and other nitrosamides. Activity in clastogenicity and SCE assays, and aberration kinetics in Chinese hamster V79-E cells. Mutation Res., 79: 239–248. Topham, J., Albanese, R., Bootman, J., Scott, D. and Tweats, D. (1983). In vivo cytogenetic assays. In: B. Dea (Ed.), Report of UKEMS Subcommittee on Guidelines for Mutagenicity Testing Part I, pp. 119–141. Tsuchimoto, T. and Matter, B.E. (1979). In vivo cytogenetic screening methods for mutagens with special reference to the micronucleus test. Archs Toxicol., 42: 239–248.

Tweats, D.J. and Gatehouse, D.G. (1988). Discussion forum: Further debate of testing strategies. Mutagenesis, 3: 95–102. Venitt, S. and Crofton-Sleigh. C. (1981). Mutagenicity of 42 coded compounds in a bacterial assay using Escherichia coli and Salmonella typhimurium. In: F.J. de Serres and J. Ashby (Eds.), Evaluation of Short-term Tests for Carcinogens. Report of the International Collaborative Program. Progress in Mutational Research, Vol 1. Elsevier, New York, pp. 351–360. Venitt, S. and Parry, J.M. (1984). Background to mutagenicity testing. In: S. Venitt and J.M. Parry, (Eds.), Mutagenicity Testing, a Practical Approach. IRL Press, Oxford, pp. 1–24. Villani, G., Boiteux, S. and Radman, M. (1978). Mechanisms of ultraviolet-induced mutagenesis: extent and fidelity of in vitro DNA synthesis on irradiated template. Proc. Natl. Acad. Sci. USA, 75: 3037–3041. Wagner, R. and Meselson, M. (1976). Repair tracts in mismatched DNA heteroduplexes. Proc. Natl. Acad. Sci. USA, 73: 4135–4139. Wahrendorf, J., Mahon, G.A.T. and Schumacher, M. (1985). A nonparametric approach to the statistical analysis of mutagenicity data. Mutation Res., 147: 5–13. Walker, G.C. (1984). Mutagenesis and inducible responses to deoxyribonucleic acid damage in Escherichia coli. Microbiol. Rev., 48: 60–93. Walker, G.C. and Dobson, P.P. (1979). Mutagenesis and repair deficiencies of Escherichia coli umu C mutants are suppressed by the plasmid pKM101. Proc. Gen. Genet., 172: 17–24. Wilcox, P., Naidoo, A., Wedd, D.J. and Gatehouse, D.G. (1990). Comparison of Salmonella typhimurium TA102 with Escherichia coli WP2 tester strains. Mutagenesis, 5: 285–291. Wolff, S. (1977a). Lesions that lead to SCEs are different from those that lead to chromosome aberrations. Mutation Res., 46: 164. Wolff, S. (1977b). In vitro inhibition of mono-oxygenase dependent reactions by organic solvents. International Conference on Industrial and Environmental Xenobiotics, Prague, Czechoslovakia. Wolff, S. and Perry, P. (1974). Differential staining of sister chromatids and the study of sister chromatid exchange without autoradiography. Chromosomes, 48: 341–353. Xu, L., Whong, W.Z. and Ong, T.M. (1984). Validation of the Salmonella (SV50)/L-arabinose-resistant forward mutation assay with 26 compounds. Mutation Res., 130: 79–86. Yahagi, T., Degawa, M., Seino, Y. et al. (1975). Mutagenicity of carcinogen azo dyes and their derivatives. Cancer Lett., 1: 91–96. Yahagi, T., Nagao, M., Seino, Y., Matsushima, T., Sugimura, T. and Okada, M. (1977). Mutagenicities of N-nitrosamines in Salmonella. Mutation Res., 48: 120–130. Yanofsky, C. (1971). Mutagenesis studies with Escherichia coli mutants with known amino acid and base-pair) changes. In: A. Hollaender (Ed.), Chemical Mutagens, Principles and Methods for Their Detection, Vol. 1. Plenum Press, New York, pp. 283–287.

13 Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

This chapter addresses a group of studies that have in common the facts that (1) they are intended to predict longer term effects that occur after repeated exposure to an agent and (2) they are tests that have historically only performed on a small subset of devices and the materials used to make them, but for which requirements are increasing. Subchronic and chronic studies for medical devices are generally in the range of being hybrids between what we are used to regarding as subchronic studies and the simple implant studies. The studies are performed using only one route—implantation—with “dose” being determined in terms of how many devices or much material are implanted. And in their simplest forms these subchronic and chronic studies are conducted as nothing more than very long implantation studies, with only the limited set of local issue tolerance indicators in the region of the implants being evaluated. It should also be kept in mind that it is frequent practice to combine such biocompatibility studies with evaluations of efficacy and/or device performance.

that are intended for continuous long-term (lifetime) use or frequent intermittent use.

REGULATORY CONSIDERATIONS Much of what is done (and how it is done) in repeat-dose studies is a response to a number of regulations. Three of these have very broad impact. These are the Good Laboratory Practices requirements, Animal Welfare Act requirements, and regulatory requirements that actually govern study design.

Good Laboratory Practices (GLPs) Since 1978, the design and conduct of preclinical safety assessment studies for pharmaceuticals in the United States (and, indeed, internationally) have been governed and significantly influenced by GLPs. Strictly speaking, these regulations cover qualifications of staff and facilities, training, record-keeping, documentation, and actions required to insure compliance with and the effectiveness of these steps. Though the initial regulations were from the U.S. FDA (FDA, 1983), they have always extended to cover studies performed overseas (FDA, 1988). Most other countries have adopted similar regulations. A discussion of these regulations is beyond the scope of the current chapter, but several aspects are central to this effort. Each technique or methodology to be employed in a study (such as animal identification, weighting and examination, blood collection, data recording, and so on) must be adequately described in standard operating procedure (SOP) before the study begins. Those who are to perform such procedures must be trained in them beforehand. The actual design of the study, including start date and how it is to be ended and analyzed, plus the principal scientists involved (particularly the study director), must be specified in a protocol that is signed before the study commences. Any changes to these features must be documented in amendments once the study has begun. It is a good idea that the pathologist who is to later perform or oversee histopathology be designated before the start of the study, and that the design be a team effort involving the best efforts of the toxicologist and pathologist.

OBJECTIVES As with any scientific study or experiment (but especially for those in safety assessment), the essential first step is to define and understand the reason(s) for the conduct of the study— that is, its objectives. There are three major (scientific) reasons for conducting subchronic studies, but a basic characteristic of all but a few subchronic studies needs to be understood. The subchronic study is (as are most other studies in whole animal toxicology) a broad screen. It is not focused on a specific end point; rather it is a broad exploration of the cumulative biological effects of the administered agent over a range of doses. So broad an exploration, in fact, that it can be called a “shotgun” study. The objectives of the typical subchronic device study fall into two categories. The first is to broadly define the toxicity of prolonged exposure to a medical device or medical device material in an animal model (most commonly, the rabbit). The second objective is one of looking forward to later studies. The subchronic study must provide sufficient information to allow a prudent setting of doses for later, longer studies (including, ultimately, carcinogenicity studies). At the same time, the subchronic study must also provide guidance for the other (than dose) design features of longer-term studies (such as what parameters to measure and when to measure them, how many animals to use and how long to conduct the study). These objectives are addressed by the usual subchronic study. Chronic studies (those that last six months or a year) may also be conducted for the above purposes but are primarily done to fulfill registration requirements for drugs

Animal Welfare Act Gone are the days when the pharmaceutical scientist could conduct whatever procedures or studies that were desired using experimental animals. The Animal Welfare Act (APHIS, 1989) (and its analogues in other countries) rightfully requires careful consideration of animal usage to ensure that research and testing uses as few animals as possible in as humane a manner as possible. As a start, all protocols must be reviewed by an institutional Animal Care and Use Committee. Such review takes time, but should not serve to 181

182

Safety Evaluation of Medical Devices

hinder good science. When designing a study or developing a new procedure or technique, the following points should be kept in mind;

Table 13.1 Test Group

1. Will the number of animals used be sufficient to provide the required data yet not constitute excessive use? (It ultimately does not reduce animal use to utilize too few animals to begin with and then have to repeat the study.) 2. Are the procedures employed the least invasive and traumatic available? This practice is not only required by regulations, but is also sound scientific practice, since any induced stress will produce a range of responses in test animals that can mask or confound the chemically induced effects.

Two–four weeks Three months Six months One year

Regulatory Requirements for Study Design The first consideration in the construction of a study is a clear statement of its objectives, which are almost always headed by meeting regulatory requirements to support device development and registration. Accordingly, the relevant regulatory requirement must be analyzed, which is complicated by the fact that new drugs are no longer developed for registration and sale in a single-market country. The expense is too great, and the potential for broad international sales too appealing. Chapter 2 should be consulted for the broad overview of such regulation.

STUDY DESIGN AND CONDUCT Animals In all but a few rare cases, medical devices are evaluated for subchronic and chronic biocompatibility in only a single species. This is most often the rabbit, though the rat, dog and hamster have also been used. The factors that should and do govern species selection are reviewed in detail in Gad and Chengelis (2006). Except in rare cases, the animals used are young, healthy adults in the logarithmic phase of their growth curve. (The FDA specifies that rodents be less than six weeks of age at the initiation of dosing.) Numbers of animals to be used in each dose group of a study are presented in Table 13.1. Though the usual practice is to use three different dose groups and at least one equalsized control group, this number is not fixed and should be viewed as a minimum (see the section on study design later in this chapter). There must be as many control animals as are in the largest-size test group. Animals are assigned to groups (test and control) by one or another form of statistical randomization. Prior to assignment, animals are evaluated for some period of time after being received in house (usually at least one week for rodents and two for nonrodents) to ensure that they are healthy and have no discernible abnormalities. The randomization is never pure; it is always “blocked” in some form or another (by initial body weight, at least) so that each group is not (statistically) significantly different from the others in terms of the “blocked” parameters (usually initial body weight). Proper facilities and care for test animals is not only a matter of regulatory compliance (and a legal requirement), but also essential for a scientifically sound and valid study. Husbandry requires clean cages of sufficient size and continuous availability of clean water and food (unless the protocol requires some restriction on their availability). Environmental conditions (temperature, humidity, and

Number of Animals for Chronic and Subchronic Study per

Study length

Rats per sex 5 20 30 50

Dogs per sex 3 6 8 10

Rabbits per sex 4 8 8 10

a Starting with 13-week studies, one should consider adding animals (particularly to the high-dose group) to allow evaluation of reversal (or progression) of effects.

light–dark cycle) must be kept within specified limits. All of these must, in turn, be detailed in the protocols of studies. The limits for these conditions are set forth in relevant NIH and USDA publications.

Setting Doses Setting of doses for longer-term toxicity studies is one of the most difficult tasks in study design. The doses administered must include one that is devoid of any adverse effect (preferably of any effect) and yet still high enough to “clear” the projected clinical dose by the traditional or regulatory safety factors (10× for rodents, 5× for nonrodents); and a second which presents the actual “dose” of clinical exposure. The resulting dose–response curve should establish: • the no adverse effect level (NOEL)—the maximum dose which an animal can tolerate over a defined period of time without showing any adverse effects; above this dose, adverse effects are observed; • and maximum implantable dose (MID)—maximum amount of implant material (dose) that a test animal can tolerate without any adverse physical or mechanical effects. Note: To avoid unnecessary morbidity in animals on a longterm test, preliminary testing may be necessary. Dose for devices and implant materials is usually evaluated in terms of exposed surface area per body mass except for the case of absorbable sutures and such where it is considered in terms of weight/body mass. Traditionally, studies include three or more dose groups to fulfill these two objectives. Based on earlier results (generally, single-dose or two-week studies), doses are selected. It is, by the way, generally an excellent idea to observe the “decade rule” in extrapolation of results from shorter to longer studies; that is, do not try to project doses for more than an order-of-magnitude-longer study (thus the traditional progression from single-dose to fourteen-day to ninety-day studies). Also, one should not allow the traditional use of three dose groups plus a control to limit designs. If there is a great deal of uncertainty, it is much cheaper in every way to use four or five dose groups in a single study than to have to repeat the entire study. Finally, remember that different doses may be appropriate for the different sexes.

PARAMETERS TO MEASURE As was stated earlier, subchronic studies are usually “shotgun” in nature, that is, they are designed to look at a very broad range of end points with the intention of screening as broadly as indications of toxicity. Meaningful findings are

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

rarely limited to a single end point—rather, what typically emerges is a pattern of findings. This broad search for components of a toxicity profile is not just a response to regulatory guidelines intended to identify potentially unsafe drugs. An understanding of all indicators of biological effect can also frequently help one to understand the relevance of findings, to establish some as unrepresentative of a risk to humans, and even to identify new therapeutic uses of an agent. Parameters of interest in the repeat-dose study can be considered as sets of measure, each with its own history, rationale, and requirements. It is critical to remember, however, that the strength of the study design as a scientific evaluation lies in the relationships and patterns of effects that are seen not in simply looking at each of these measures (or groups) as independent findings, but rather as integrated profiles of biological effects.

Body Weight Body weight (and the associated calculated parameter of body weight gain) is a nonspecific, broad screen for adverse systemic toxicity. Animals are initially assigned to groups based on a randomization scheme that includes having each group vary insignificantly from one another in terms of body weight. Weights are measured prior to the initial dose, then typically one, three, five, seven, eleven, and fourteen days thereafter. The frequency of measurement of weights goes down as the study proceeds—after two weeks, weighting is typically weekly through six weeks, then every other week through three months, and monthly thereafter. Because the animals used in these studies are young adults in the early log phase of their growth, decreases in the rate of gain relative to control animals is a very sensitive (albeit nonspecific) indicator of systemic toxicity.

Food Consumption Food consumption is typically measured with one or two uses in mind. First, it may be explanatory in the interpretation of reductions (either absolute or relative) in body weight. In cases where administration of the test compound is via diet, it is essential to be able to adjust dietary content so as to accurately maintain dose levels. Additionally, the actual parameter itself is a broad and nonspecific indicator of systemic toxicity. Food consumption is usually measured over a period of several days, first weekly and then on a once-a-month basis. Water consumption, which is also sometimes measured, is similar in interpretation and use.

Clinical Signs Clinical signs are generally vastly underrated in value, probably because insufficient attention is paid to care in their collection. Two separate levels of data collection are actually involved here. The first is the morbidity and mortality observation, which is made twice a day. This generally consists of a simple cage-side visual assessment of each animal to determine if it is still alive, and, if so, whether it appears in good (or at least stable) health. Historically, this regulatorily required observation was intended to ensure that tissues from intoxicated animals were not lost for meaningful histopathologic evaluation due to autolysis (Fox, J.C., 1997; Arnold et al., 1990). The second level of clinical observation is the detailed hands-on examination analogous to the human physical examination. It is usually performed against a checklist (see Gad and Chengelis, 1998, for an example), and evaluation is of the incidence of observations of a particular type in

Table 13.2 Test Group

183

Number of Animals for Chronic and Subchronic Study per

Clinical chemistry

Hematology

Urinalysis

Albumin Alkaline phosphatase (ALP) Blood urea nitrogen (BUN) Calcium Chloride Creatine Creatine phosphokinase (CPK) Direct bilirubin Gamma glutamyl transferase (GGT) Globulin Glucose Lactic dehydrogenase (LDH) Phosphorus Potassium Serum glutamic-oxaloacetic transaminase (SGOT) Serum glutamic-pyruvic transaminase (SGPT) Sodium Total bilirubin Total cholesterol Total protein Triglycerides

Erythrocyte count (RBC) Hemoglobin (HGB) Hematocrit (HCT) Mean corpuscular hemoglobin (MCH) Mean corpuscular volume (MCV) Platelet count Prothrombin time Reticulocyte count White cell count (WBC) White cell differential count

Chloride Bilirubin Glucose Ketone Osmolality Occult blood pH Phosphorus Potassium Protein Sodium Specific gravity Volume

a group of treated animals compared to controls. Observations range from being indicative of nonspecific systemic toxicity to fairly specific indicators of target organ toxicity. These more detailed observations are typically taken after the first week of a study and on a monthly basis thereafter. Ophthalmologic examinations are typically made immediately prior to initiation of a study (and thus serve to screen out animals with preexisting conditions) and toward the end of a study. Particularly when the agent under investigation either targets or acts via a mechanism likely to have a primary effect on a certain organ for which functional measures are available, an extra set of measurements of functional performance should be considered. The organs or organ systems that are usually of particular concern are the kidneys, liver, cardiovascular, nervous and immune. Special measures (such as creatinine clearance as a measure of renal function) are combined with other data already collected (organ weights, histopathology, clinical pathology, etc.,) to provide a focused “special” investigation or evaluation of adverse effects on the target organ system of concern.

Clinical Pathology Clinical pathology covers a number of biochemical and morphological evaluations based on invasive and noninvasive sampling of fluids from animals that are made periodically during the course of a subchronic study. These evaluations are sometimes labeled as clinical (as opposed to anatomical) pathology determinations. Table 13.2 presents a summary of the parameters measured under the headings of clinical chemistry, hematology, and urinalysis, using samples of blood and urine collected at predetermined intervals during the study. Conventionally, these intervals are typically at three points evenly spaced over the course of the study, with the

184

Safety Evaluation of Medical Devices

Table 13.3

Association of Changes in Biochemical Parameters with Actions at Particular Target Organs

Organ system Parameter

Blood

Heart

Lung

Albumin

Kidney

Liver

Ø

Ø ≠

ALP (alkaline (phosphatase) Bilirubin (total)

Bone

BUN (Blood urea (nitrogen)

≠

Calcium

≠ ≠

Ø

Ø

≠ ≠

Glucose

≠

GGT (gamma glutamyl transferase

≠

≠ ≠

Protein (total)

SGOT (serum glutamicoxaloacetic transaminase); also called AST (aspirate amino transferase) SGPT (serum glutamicpyruvic transaminase; also called ALT (alanide amino transferase) SDH (sorbitol dehydrogenase)

≠

≠

Creatine

HBDH (hydroxybutyric dehydrogenase) LDH (lactic dehydrogenase)

Pancreas

≠

≠

Cholinesterase CPK (creatinine (phosphokinase)

≠

Intestine

≠

≠

≠

≠

≠

≠

≠

≠

≠

≠

Produced by the liver. Very significant reductions indicate extensive liver damage Elevations usually associated with cholestasis. Bone alkaline phosphatase tends to be higher in young animals. Usually elevated due to cholestasis either due to obstruction or hepatopathy. Estimates blood-filtering capacity of the kidneys. Doesn’t become significantly elevated until kidney function is reduced 60–75%. Can be life threatening and result in acute death. Found in plasma, brain and RBC. Most often elevated due to skeletal muscle damage but can also be produced by cardiac muscle damage. Can be more sensitive than histopathology. Also estimates blood-filtering capacity of kidney as BUN does. More specific than BUN. Alterations other than those associated with stress are uncommon and reflect an effect on the pancreatic islets or anorexia. Elevated in cholestasis. This is a microsomal enzyme and levels often increase in response to microsomal enzyme induction. Most prominent in cardiac muscle tissue. Increase usually due to skeletal muscle, cardiac muscle, and liver damage. Not very specific unless isozymes are evaluated. Absolute alterations are usually associated with decreased production (liver) or increased loss (kidney). Present in skeletal muscle and heart and most commonly associated with damage to these.

Elevations usually associated with hepatic damage or disease.

Liver enzyme which can be quite sensitive but is fairly unstable. Samples should be processed as soon as possible.

≠ or Ø

first being one month after study initiation and the last being immediately prior to termination of the test animals. For a three-month study, this means that samples of blood and urine would be collected at one, two, and three months after study initiation (i.e., after the first day of dosing of the animals). There are some implications of these sampling plans that should be considered when the data are being interpreted. Many of the clinical chemistry (and some of the hematologic) markers are really the result of organ system damage that may be transient in nature (see Table 13.3 for a summary of interpretations of clinical chemistry findings and Table 13.4 for a similar summary for hematological findings). The samples on which analysis is performed are from fixed points in time, which may miss transient changes (typically, increases) in some enzyme levels.

Notes

HISTOPATHOLOGY Histopathology is generally considered the single most significant portion of data to come out of a repeat-dose toxicity study. It actually consists of three related sets of data (gross pathology observations, organ weights, and microscopic pathology) that are collected during the termination of the study animals. At the end of the study, a number of tissues are collected during termination of all surviving animals (test and control). Organ weights and terminal body weights are recorded at study termination, so that absolute and relative (to body weight) values can be statistically evaluated. These tissues, along with the organs for which weights are determined, are listed in Table 13.4. All tissues collected

Chapter 13:

Table 13.4 Changes

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

Some Probable Conditions Affecting Hematological

Parameter

Elevation

Depression

Red blood cells

1. 2. 3. 4.

1. Anemias a) Blood loss b) Hemolysis c) Low RBC production

Hematocrit

1. Increased RBC 2. Stress 3. Shock (a) Trauma (b) Surgery 4. Polycythemia

1. Anemias 2. Pregnancy 3. Excessive hydration

Hemoglobin

1. Polycythemia (increase in production of RBC)

1. Anemias 2. Lead poisonings

Vascular shock Excessive diuresis Chronic hypoxia Hyperadrenocorticism

Mean cell volume 1. Anemias 2. B-12 deficiency

1. Iron deficiency

Mean corpuscular 1. Reticulocytosis hemoglobin

1. Iron deficiency

White blood cells

1. Bone marrow depression 2. Cancer chemotherapy 3. Chemical intoxication 4. Splenic disorders

1. Bacterial infections 2. Bone marrow stimulation

1. 2. 3. 4. 5. 6.

Lymphocytes

1. Leukemia 2. Malnutrition 3. Viral infections

Monocytes

1. Protozoal infections

Eosinophils

1. 2. 3. 4.

Basophils

1. Lead poisoning

Adrenalsa Body and cervix Brain, all three levelsa Cervical lymph nodes Cervical spinal cord Duodenum Esophagogastric junction Esophagus Eyes with optic nerves Femur with marrow Heart Ileum Kidneysa Large bowel Larynx with thyroid and parathyroid Livera Lungsa

Mainstream bronchi Major salivary gland Mesenteric lymph nodes Ovaries and tubes Pancreas Pituitary Prostate Skeletal muscle from proximal hind limb Spleena Sternbrae with marrow Stomach Testes with epididymidesa Thymus and mediastinal contentsa Thyroid with parathyroida Trachea Urinary bladder Uterus including horns

Organs to be weighed.

Acute bacterial infections 1. Viral infections Tissue necrosis Strenuous exercise Convulsions Tachycardia Acute hemorrhage

Neutrophilis

Tissues for Histopathology

a

1. Bone marrow depression 2. Immune disorder

Platelets

Table 13.5

185

Allergy Irradiation Pernicious anemia Parasitism

are typically processed for microscopic observation, but only those from the high-dose and control groups are necessarily evaluated microscopically. If a target organ is discovered in the high-dose group, then successively lower-dose groups are examined until a ‘clean” (devoid of effect) level is discovered (Haschek and Rousseaup, 1991). In theory, all microscopic evaluations should be performed blind (without the pathologist knowing from which dose group a particular animal came), but this is difficult to do in practice and such an approach frequently degrades the quality of the evaluation. Like all the other portions of data in the study, proper evaluation benefits from having access to all data that addresses the relevance, severity, timing, and potential mechanisms of a specific toxicity. Blind examination is best applied in peer review or consultations on specific findings.

In addition to the “standard” set of tissues specified in Table 13.5, observations during the course of the study or in other or previous studies may dictate that additional tissues be collected or special examinations (e.g., special stains, polarized light or electron microscopy, immunocytochemistry, or quantitative morphometry) be undertaken to evaluate the relevance of, or understand the mechanisms underlying, certain observations. Histopathology testing is a terminal procedure, and, therefore, sampling of any single animal is a one-time event (except in the case of a tissue collected by biopsy). Because it is a regulatory requirement that the tissues from a basic number of animals be examined at the stated end of the study, an assessment of effects at any other time course (most commonly, to investigate recovery from an effect found at study termination) requires that satellite groups of animals be incorporated into the study at start-up. Such animals are randomly assigned at the beginning of the study, and otherwise treated exactly the same as the equivalent treatment (or control) animals. For devices, the tissue list may either be limited to or augmented by the addition of specific examination of the tissues at the site(s) of implantation as follows: • section, stain and examine implant sites and other sites showing gross pathology • score lesions on numerical scale from normal to extreme (1 to 5) • criteria for implant site acute inflammatory response subacute/chronic response inflammatory cells (PMNs) mononuclear necrosis inflammatory cells hemorrhage (lymphocytes, fibrin or serum macrophages, eosinophils, plasma cells) epithelial cells giant cells fibroblasts • Components of sub-acute or chronic response S mononuclear inflammatory cells (lymphocytes, macrophages, plasma cells) S epithelioid or giant cells S fibroplasia or fibrosis

186

Safety Evaluation of Medical Devices

• measure width of reactive zone • score on scale of 0 (not present) to 5 (extreme) • components of acute inflammatory response S inflammatory cells (polymorphonuclear leukocytes) S necrosis S hemorrhage S fibrin/serum.

STUDY INTERPRETATION AND REPORTING For a successful repeat-dose study, the bottom line is the clear demonstration of a no-effect level, characterization of a toxicity profile (providing guidance for any clinical studies), and at least a basic understanding of the mechanisms involved in any identified pathogenesis. The report that is produced as a result of the study should clearly communicate these points—along with the study design and experimental procedures, summarized data, and their statistical analysis—and it should be GLP compliant, suitable for FDA submission format. Interpretation of the results of a study should be truly scientific and integrative. It is elementary to have the report state only each statistically and biologically significant finding in an orderly manner. The meaning and significance of each in relation to other findings, as well as the relevance to potential human effects, must be evaluated and addressed. The author of the report should insure that it is accurate and complete, but also that it clearly tells a story and concludes with the relevant (to clinical development) findings.

REPRODUCTIVE AND DEVELOPMENTAL TOXICITY Reproductive toxicity tests for medical devices should normally be considered for the following: (a) intrauterine devices (IUDs), or any other long-term contact devices likely to come into direct contact with reproductive tissues or the embryo/fetus; (b) energy-depositing devices; (c) resorbable or leachable materials and devices. There is no need for the testing of resorbable devices or devices containing leachable moieties where there is adequate and reassuring data from absorption, metabolism, distribution, and on the reproductive toxicity of all major components identified in extracts. Individual compounds known to cause reproductive toxicity should not be present as significant components of extracts of materials or devices Such tests are intended and should be adequate to evaluate the potential effects of devices, materials and/or extracts on reproductive function, embryonic development (developmental toxicity tenatagenicity), and prenatal and early postnatal development.

Introduction The goal of testing the developmental and reproductive toxicity of candidate devices and device material in laboratory animals is to predict which agent would adversely affect pregnancy and reproduction in humans. The types of developmental and reproductive toxicity studies performed prior to 1993 and the methods used have been extensively documented (see Palmer, 1981; Christian, 1983; Heinrichs, 1985; Heywood and James, 1985;

Persuad, 1985; Schardein, 1988; Tyl, 1988; Christian and Hoberman, 1989; Khera, 1985; Manson and Kang, 1989, Hood, 2007). Since June 20, 1979, the FDA has required that these studies be conducted according to Good Laboratory Practice (GLP) regulations (see FDA, 1978, 1987). The conduct of these studies had been complicated by the need to satisfy worldwide regulatory guidelines that varied from country to country. As a result, studies were conducted for regulatory purposes that, from a scientific viewpoint, were redundant, superfluous, and/or unnecessarily complex. This situation was changed in 1993 when the International Conference on Harmonization of Technical Requirements for the Registration of Pharmaceuticals for Human Use (ICH) standardized worldwide requirements in the guideline “Detection of Toxicity to Reproduction for Medicinal Products.” ISO is likewise working on guidelines for medical devices.

ICH Study Designs The new ICH guideline allows for various combinations of studies (note: the complete guideline is included in the appendix of this volume). The studies conducted must include evaluation of the following components: 1. Male and female fertility and early embryonic development to implantation 2. Embryo–fetal development 3. Pre- and postnatal development including maternal function. These components would normally be evaluated in a rodent species (preferably rats) and, in addition, embryo–fetal development would be evaluated in a second species, typically the rabbit. The “most probable option” was considered in the ICH guideline to be the case where three rodent studies would be conducted that separately addressed each of the components listed above. These study designs are described below. The day of insemination or detection of evidence of mating is considered Day 0 of gestation and the day of birth is considered postpartum and postnatal Day 0.

Male and Female Fertility and Early Embryonic Development to Implantation The purpose of this component is to assess the effects that result from treatment during maturation of gametes, during cohabitation, and, in females, during gestation up through the time of embryo implantation (typically last dose on Day 6 of gestation). Assuming that the findings from a toxicity study of at least one month in duration do not contraindicate, the treatment period begins in males four weeks before male/female cohabitation and, in females, two weeks prior to cohabitation. A group size of 16–24 litters would generally be considered acceptable (the author recommends mating 26 males with 26 females). Minimal in-life observations include: 1. 2. 3. 4. 5.

Clinical signs and mortality daily Body weight twice weekly Food consumption weekly Vaginal cytology daily during cohabitation Valuable target effects seen in previous toxicity studies.

Females are sacrificed after the middle of the gestation period. Males are sacrificed at any time after the end of the cohabitation period, but it is generally advisable to retain the males until after the outcome of the first mating is known, to ensure that a repeat cohabitation with untreated females will not be needed to determine if an observed effect on mating

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

performance is a male effect. Males are treated until termination. Terminal examination of adults includes; 1. Necropsy 2. Preservation of organs with gross changes and sufficient control organs for comparison 3. Preservation of testes, epididymides, ovaries, and uteri 4. Sperm count and sperm viability 5. Count of corpora lutea and implantation sites 6. Count of live and dead conceptuses

187

A minimum of 50% of fetuses are to be examined for visceral alterations and a minimum of 50% for skeletal abnormalities. When a fresh microdissection technique is being used for the visceral examination of rabbit fetuses, all fetuses should be examined for both visceral and skeletal abnormalities.

Pre- and Postnatal Development

Among the study designs conducted before the ICH guidelines, the Segment I fertility study conducted according to Japanese guidelines is most similar to this ICH study design. The major differences are the shortening of the treatment period of males prior to cohabitation from the duration of spermatogenesis (60–80 days) to four weeks and the addition of sperm evaluation. The justifications given for shortening the treatment period of males are:

The purpose of this component is to detect effects of treatment from implantation through lactation on the pregnant and lactating female and on the development of the conceptus and offspring through sexual maturity. The study design is similar to the previous Segment III study design except that dosing begins on Day 6 of gestation instead of Day 15. A group size of 16–24 litters would generally be considered acceptable (the author recommends 25 mated females). Minimal in-life observations for parental (F0 generation) females include:

1. Careful organ weight and histopathological evaluation of testes in general toxicity studies will detect most testicular toxins. 2. Fertility is an insensitive measure of testicular effects. 3. Compounds known to affect spermatogenesis generally exert their effects during the first four weeks of treatment.

1. 2. 3. 4. 5. 6.

Sperm counts can be performed with sperm from either the testis or the epididymis. Sperm motility is commonly being treated as a measure of sperm viability. The addition of sperm evaluation greatly increases the sensitivity of the study to detect effects on sperm maturation and the new study design will likely detect more male effects than the previous design, even though the treatment period number has been shortened.

Parental females are sacrificed after weaning of the F1 generation. The age of sacrifice of the F1 generation animals is not specified in the ICH guideline and varies among laboratories. Typically, they are sacrificed intermittently with some laboratories reducing litter size on postnatal Day 0, 3, or 4, on postnatal Day 21 or at weaning, at male/female cohabitation to produce an F2 generation, and the terminal sacrifice, after production of the F2 generation. Terminal examinations for maternal animals and offspring include:

Embryo–Fetal Development The purpose of this component is to detect anatomical effects on the developing conceptus by treating during the period of organogenesis from implantation to closure of the secondary palate. The study design is very similar to the previous Segment II developmental toxicity study. A group size of 16–24 litters would generally be considered acceptable. The author recommends the following:

Treatment period (gestational days) Group size (mated or inseminated)

Rat

Rabbit

Mouse

6–17 25

6–18 20

6–15 25

Minimal in-life observations include: 1. 2. 3. 4.

Clinical signs and mortality daily Body weight twice weekly Food consumption weekly Valuable target effect seen in previous toxicity studies.

Females are sacrificed at the end of the gestation period, about one day prior to parturition (Day 20 or 21 for rats, Day 28 or 29 for rabbits, and Day 17 or 18 for mice). Terminal examinations include: 1. Necropsy 2. Preservation of organs with gross changes and sufficient control organs for comparison 3. Count of corporea lutea and live and dead implantations 4. Fetal body weight 5. External, visceral, and skeletal examinations of fetuses 6. Gross evaluation of placenta.

Clinical signs and mortality daily Body weight twice weekly Food consumption weekly Valuable target effects seen in previous toxicity studies Length of gestation Parturition

1. Necropsy of all parental and F1 adults 2. Preservation of organs with gross changes and sufficient control organs for comparison 3. Count of implantations. Additional observations of the F1 generation include: 1. 2. 3. 4.

Abnormalities Live and dead offspring at birth Body weight at birth Pre- and postnatal survival, growth, maturation, and fertility 5. Physical development including vaginal opening and preputial separation 6. Sensory function, reflexes, motor activity, learning, and memory.

Single-study and Two-study Designs for Rodents Except for the embryo–fetal development component in rabbits, the components described above can be combined into fewer, larger studies instead of conducting each component separately. Acceptable alternatives include the “single-study design” and “two-study design.” In the “single-study design,” all of the above components are combined into one study. The dosing period, extending from before mating through lactation, is a combination of that for the fertility study together with that for the pre- and postnatal development study. Subgroups of animals are terminated at the end of gestation for fetal examination. There are a variety of possible “two-study designs.” One is to conduct the single study described above except

188

Safety Evaluation of Medical Devices

that, instead of having subgroups for fetal examination, a separate embryo–fetal development study in rodents is conducted. Another two-study design consists of combining the embryo–fetal development study with the pre- and postnatal development study such that the two studies to be conducted would be (1) the fertility study and (2) the pre- and postnatal development study with subgroups terminated at the end of gestation for fetal examination. A third two-study design is to combine the fertility study with the embryo–fetal development study. In the first study, treatment would extend through the end of organogenesis and then, at termination at the end of gestation, there would be a complete fetal examination. The second study would be the pre- and postnatal development study. For all the options described above, effects on male and female fertility can be evaluated separately by conducting separate studies in which only one sex is treated. The treatment periods are the same, but the treated animals are cohabited with untreated animals of the opposite sex. In the male fertility study, the untreated females are terminated after the middle of gestation and terminal observations include embryo survival and possibly external examination of fetuses (if terminated at the end of gestation; see Tanimura, 1990). The advantage of conducting separate male and female studies is that, if there are effects, it is clear which sex was affected by treatment. Often when effects are seen in a combined male and female study, additional work is required to resolve which sex was affected. Either a second cohabitation of the treated males with untreated females is added or studies with only one sex treated must then be conducted. With the possible exception of combining the female fertility component with the embryo–fetal development component, the combined-study approach will likely not be used often. The female fertility and embryo–fetal development components are needed to support clinical trials in women of childbearing potential in most countries and thus will be conducted early in the development of a drug. However, since the pre- and postnatal development component is not routinely required for clinical studies of women of childbearing potential and represents a large commitment of resources, it will not generally be conducted until late in the drug development process.

DOSE AND SAMPLE PREPARATION In the case of energy-depositing devices, whole-body irradiation of the animals with a multiple of the dose to be expected in humans should be applied. When possible, IUDs, resorbable devices, or devices containing leachable moieties shall be tested in their “ready-to-use” form. Otherwise a suitably formed implant shall be made of the test material. The maximum implantable dose (MID) of a material or device should be applied. Where possible this dose should be expressed as a multiple of the worst case human exposure (in milligrams per kilogram).

METHODOLOGICAL ISSUES Control of Bias An important element to consider when designing developmental and reproductive toxicity studies, as in any study, is the control of bias. For example, animals should be assigned to groups randomly and preferably balanced by body weight. This can be accomplished by first ranking the animals in

order of body weight and then, starting with the lightest or heaviest, assigning by rank to groups based on a list of sets or random permutations of numbers (e.g., 1, 2, 3, and 4 if there are four groups, where 1 represents the control group, 2 represents the low-dose group, etc.). Housing of the treatment groups should also be unbiased. This can be by “Latin square” design where each block of four cages (if there are four groups) includes an animal from each group. It is often an acceptable compromise to have animals from different groups in alternating vertical columns with all the animals in a column from the same group. This provides equal vertical balancing for all groups. The order of sacrifice on the day of cesarean sectioning should be balanced by group (again using random permutations) since fetuses continue to grow during the day and an unbalanced time of sacrifice would bias fetal weights, particularly for rodents. Alternatively, all animals can be killed at about the same time in the morning and the fetuses stored for examination later the same day. Fetal examinations should be conducted blind, that is, without knowledge of the treatment group.

Diet It is known that rodents require a diet relatively rich in protein and fats for successful reproduction (Zeman, 1967; Chow and Rider, 1973; Turner, 1973; Mulay et al., 1982). Consequently, rodents are fed high-protein, high-fat diets ad lib for reproductive toxicity studies and also generally as a maintenance diet for all toxicity studies. Female rats fed in this manner being to show decreases in fertility, litter size, and the incidence of normal estrus cycling at the age of six months (Matt et al., 1986, 1987). The disadvantage of this feeding practice is that the animals more quickly acquire age-related diseases and sexual dysfunctions and die sooner than if they were fed a restricted amount of calories (for review, see Weindruch and Walford, 1988). In relatively short-term studies (such as the standard ICH studies), this rapid aging does not present a problem. However, for male breeding colonies or multigeneration studies with multiple litters per generation, it could be advantageous to restrict caloric intake, at least when the animals are not being bred. Restriction of food intake to achieve adverse effect on male rat reproduction (Chapin et al., 1991), although it does affect reproduction in mice (Gulati et al., 1991) and female rats (Chapin et al., 1991). Thus, it is desirable to improve the assessment of toxicity in range-finding studies in pregnant animals. Complete histopathologic examination is too impractical. However, it is often feasible to perform hematologic and serum biochemical analyses that can significantly increase the chances of detecting significant toxicity and provide important information for selecting an appropriate highest dosage level for the embryo–fetal developmental toxicity study. Body weight effects most often provided the basis for selection of dosage levels in the Segment II study. However, there have been cases where clinical pathology was or would have been useful to justify dosage selection. For example, the nonsteroidal anti-inflammatory drug diflunisal caused a decrease in erythrocyte count from 6.0 (million/mm3) to 2.9 at a dosage level (40 mg kg−1 day−1) that caused only a 1% decrease in body weight in pregnant rabbits. The severe hemolytic anemia caused by this excessively high dosage level in turn caused secondary axial skeletal malformations in the fetuses (Clark et al., 1984). Also, the angiotensinconverting enzyme (ACE) inhibitor enalapril caused an

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

increase in serum urea nitrogen from 16 to 46 mg dL−1 (highest value = 117) at a dosage level (10 mg kg−1 day−1) that had no apparent effect on body weight but caused a significant (p < 0.05) increase in resorptions. Serum urea nitrogen concentration was used to select dosage levels for a subsequent ACE inhibitor, lisinopril. The routine use of clinical pathology in range-finding studies has previously been proposed (Wise et al., 1988). The animals can be bled on the day after the last dose or sooner to detect transient effects or to allow an evaluation of the data prior to cesarean section.

GRAVID UTERINE WEIGHTS Effects of treatment on maternal body weight gain are commonly evaluated as indicators of maternal toxicity. However, maternal body weight gain is influenced by fetal parameters such as live fetuses per litter and fetal body weight. Thus, effects indicative of developmental toxicity could contribute to decreased maternal body weight gain and confound the interpretation of maternal toxicity. In addition, other maternal but pregnancy-related parameters, such as volume of intrauterine fluid, could be affected by treatment and contribute to effects on overall body weight gain. In an attempt to correct this complication, some laboratories weigh the gravid uterus at cesarean section and then subtract the weight of the gravid uterus from the body weight gain to obtain an adjusted weight gain that is more purely maternal. This adjustment may be imprecise but not inappropriate for rats for which gravid uterine weight is correlated with and generally substantially less than maternal body weight change during gestation. However, the subtraction of gravid uterine weight from maternal weight gain is an over-adjustment for rabbits. The maternal body weight gain of rabbits during gestation is generally less than the weight of the gravid uterus. Moreover, gravid uterine weight is correlated with maternal body weight change in some but not all studies. Thus, subtracting the gravid uterine weight from the maternal weight gain is not always appropriate. A preferred method for adjusting maternal body weight gain for possible developmental effects is to test and, if appropriate, use gravid uterine weight as a covariate (J. Antonello, personal communication, 1990). This method can be used for both rats and rabbits and for body weight change intervals in addition to those ending at study termination. Alternatively, to avoid weighing the uterus (or if the analysis is being performed retrospectively and uterine weights are unavailable) or if a more purely fetal adjustment is desired, one can use the sum of the live fetal weights within the litter (total live fetal weight) as the covariate instead of gravid uterine weight. As one would expect, total live fetal weight is very highly correlated with gravid uterine weight in control animals (r = 0.99 in control rats and 0.95 in control rabbits; J. Antonello, personal communication, 1990). Thus, in general, using either gravid uterine weight or total live fetal weight as the covariate will yield similar results. However, if treatment was to have an effect on gravid uterine weight that was not reflected in total live weight (e.g., if the volume of amniotic, extracoelomic, or intrauterine fluid was affected), then total live fetal weight may not be highly correlated with gravid uterine weight and, hence, not interchangeable as a covariate. In that case, only weighing the gravid uterus would allow the detection of these effects not revealed by total live fetal weight.

189

IMPLANT COUNTS AND DETERMINATION OF PREGNANCY Two observations suggest that the remnants of embryos that die soon after implantation are not apparent at gross examination of the uterus near term. First, embryos that were observed to be resorbing at laparotomy early in gestation left no readily visible trace near term (Staples, 1971). Second, occult implantation sites can be revealed near term by staining the uterus with ammonium sulfide or sodium hydroxide (Salewski, 1964; Yamada et al., 1988). It is not known if the uterine staining techniques reveal all implantation sites. It is clear, though, that when uterine staining techniques are not used, very early resorptions may not be included in what is termed the “resorption rate” but instead may contribute to the apparent “preimplantation loss” or, if no implantation sites were detected, the rate of “nonpregnant” animals. In normal circumstances, probably very few implantation sites are not detected without staining. However, cases have occurred in which probable treatment effects were detected only as a result of uterine staining.

Fetal Examinations Many fetal anomalies, such as cleft palate, exencephaly, ectrodactyly, and missing vertebra, are discrete and therefore easy to recognize objectively. Some anatomical structures, though, occur along a continuous gradation of size and shape and are only considered anomalous if the deviation from the typical exceeds a somewhat arbitrarily selected threshold. These anomalies are observed in all examination types and include, for example, micrognathia, reduced gallbladder, enlarged heart, distended ureter, wavy rib, and incomplete ossification at many sites. In many cases, it cannot be said with certainty whether a specific degree of variation from normal would have resulted in an adverse consequence to the animal and should therefore be considered abnormal. In the absence of certainty about outcome, the best approach is to uniformly apply a single criterion within a study (and preferably among studies) so that all treatment groups are examined consistently. The subjectivity (and hence fetus-to-fetus variability) of the examination can be minimized by having the criteria be as clear and objective as possible. For example, when examining for incompletely ossified thoracis centra or supraoccipitals, it can be required that the ossification pattern be absent (unossified), unilateral, or bipartite (which are objective observations) before recording and observation. Subjective criteria such as being dumbbell- or butterfly-shaped would not be applied. Additional comments about specific examination types follow.

Examination of External Genitalia One aspect of external anatomy that is largely overlooked in the examination of offspring exposed in utero to test agents is the external genitalia, even though major malformations can occur in those structures. For example, hypospadias is a malformation in the male in which the urethra opens on the underside of the penis or on the perineum. Hypospadias can occur in the male rat following in utero exposure to antiandrogens, testosterone synthesis inhibitors (e.g., Bloch et al., 1971), or finasteride, a 5α-reductase inhibitor (Clark et al., 1990b). However, it is impractical to detect hypospadias in fetuses or young pups. Although the genital tubercle of the normal male rat fetus is grossly distinguishable from that of the normal female as early as Day 21 of gestation (the female has a groove on the ventral side), the difference is very

190

Safety Evaluation of Medical Devices

subtle, and partial feminization of the male genital tubercle would be very difficult to ascertain. Routine histological examination is obviously too labor intensive to be considered. Hypospadias can be readily determined, though, by expressing and examining the penis of the adult. Thus it is recommended that adult F1 males be examined for hypospadias. If the timing of the separation of the balano-preputial membrane is being included in the pre- and postnatal development study as a developmental sign (see Korenbrot et al., 1977), the examination of the penis for hypospadias can be conducted at the same time. The critical period for the induction of hypospadias by finasteride in rats is Day 16 to 17 of gestation (Clark et al., 1990a). It is unlikely that other agents would have a much earlier critical period since testosterone synthesis, which is required for the development of the penile urethra, begins in the rat on Day 15 of gestation (Habert and Pcion, 1984). Thus, if treatment in the embryo-fetal development study terminates on Day 15 of gestation (as is done in some laboratories), it is doubtful that hypospadias could be induced. However, hypospadias could be induced in the prepostnatal development study. Since the formation of the penile urethra in the rat is not completed until Day 21 of gestation (Anderson and Clark, 1990), it could be argued that “major organogenesis” continues until that time. One parameter that is readily and commonly measured as an indicator of effects on differentiation of the external genitalia in rodent fetuses is the sexually dimorphic distance between the anus and the genital tubercle (anogenital distance). However, it should not be assumed that anogenital distance is synonymous with hypospadias, since effects on anogenital distance are not necessarily predictive of hypospadias. Finasteride caused both hypospadias and the decreased anogenital distance in male offspring but with very different dose–response relationships and only a slight tendency for animals with hypospadias to have a shorter anogenital distance (Clark et al., 1990b). Also, the effects on anogenital distance in male rat fetuses on Day 20 of gestation did not affect the development of the genital tubercle and did not cause hypospadias (Wise et al., 1990a). Thus, decreased anogenital distance per se does not necessarily indicate a serious congenital anomaly. When evaluating effects of treatment on fetal anogenital distance, it is obviously important to correct for effects on fetal weight. One approach is to calculate “relative” anogenital distance, the ratio between anogenital distance and another linear measure, for example, biparietal diameter (head width). The cube root of fetal weight simulates a linear measure (Wise et al., 1990b) and can also be used to normalize anogenital distance. Another approach is to compare the anogenital distance in a weight-reduced treatment group to that in a weight-matched control group at a younger age.

Visceral Fetal Examinations The examination of the abdominal and thoracic viscera of fetuses is performed either fresh without fixation (“Staples technique”) or after Bouin’s fixation by making freehand razor blade sections (“Wilson’s technique”; Wilson, 1965). Both techniques have advantages. The fresh examination technique, which may require less training to become thoroughly proficient, provides a more easily interpreted view of heart anomalies. The examination must be performed on the day the dam is terminated, though, so having a large number of litters to examine in one day requires that a large team of workers be committed to the task.

With both techniques, the heads of one-half of the fetuses can be fixed in Bouin’s fixative for subsequent freehand sectioning and examination. A common artifact induced by fixation in rabbit fetal heads is retinal folding.

Skeletal Fetal Examination There is variability in the development of the fetal skeleton, including numbers of vertebrae and ribs, patterns of sternebral ossification, alignment of ribs with sternebrae, and alignment of ilia with lumbar and sacral vertebrae. There is also extensive plasticity in the development of the skeleton beyond the fetal stage. For example, it is known that markedly wavy ribs in fetuses can resolve so that the ribs in the adult are normal (Saegusa et al., 1980) and supernumerary ribs can be resorbed (Wickramaratne, 1988). This variability and plasticity complicates the classification of anomalies as true malformations as opposed to variations of normal. There is no unanimity on terminology, but in general, a variation tends to be an alteration that occurs at a relatively high spontaneous incidence (>1%), is often reversible, and has little or no adverse consequence for the animal. When tabulating and interpreting fetal skeleton data, a distinction is made between alterations in the pattern of development and simple delays in development that are considered to be less serious. A delay in skeletal development is usually apparent as a delay in ossification, as evidenced by an increased incidence of specific, incompletely ossified sites or decreases in counts of ossified bones in specific regions (e.g., sacrocaudal vertebrae). These delays are normally associated with decreases in fetal weight and commonly occur at dosage levels of the test agent that also cause decreased maternal body weight gain. When determining the criteria for recording skeletal alterations, particularly sites of incomplete ossification, it is legitimate to consider the resulting incidences. For example, including an unossified fifth sternebra in the criteria for recording incomplete sternebra ossification may increase the control incidence to a very high proportion (over 95%) of fetuses affected, which would then reduce the sensitivity for detecting treatment effects. The additional effort expended in recording the extra observations due to sternebra 5 would be wasted. In addition, recording high incidences of complete ossification at many sites is not worth the effort involved. The ossification at various sites is highly correlated, so recording at multiple sites is redundant. In some cases, the incidences can be reduced to reasonable levels (1–20% of control fetuses) and the criteria simultaneously made more objective by requiring that the bone be entirely unossified before recording.

DEVELOPMENTAL SIGNS The postnatal evaluation of F1 pups includes the observation of developmental signs in two or more pups per sex per litter. In general, the acquisition of these developmental landmarks, including anatomical changes (e.g., ear pinna detachment, incisor eruption, hair growth, and eye opening) and reflexes (negative geotaxis, surface righting, and free-fall righting), are highly correlated with body weight but as indicators of developmental toxicity they are not as sensitive as body weight (Lochry et al., 1984; Lochry, 1987) and thus have minimal value. Possible exceptions to this generality are the ontogeny of the auditory startle reflex and the markers of

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

sexual maturation (vaginal patency, testes descent, and balano-preputial separation in males). The examinations for developmental signs can be performed daily, starting before and continuing until criterion is achieved. Alternatively, the examinations can be conducted on animals on preselected postnatal days. For example, three days can be selected—one when only a few control animals will achieve criterion, one when approximately one half of control animals will achieve criterion, and one when nearly all control animals will achieve criterion. The latter method has the advantages of requiring less time and assuring that all examined animals have the same testing experience (so as not to bias subsequent tests). The separation of the balano-preputial membrane of the penis (occurring at postnatal Week 6 to 7; Korenbrot et al., 1977) is becoming the preferred landmark of sexual maturation in males. The timing of testes descent is more variable and very dependent on the achievement criteria used. Another advantage of determining the time of balano-preputial separation is that anomalies of the penis may be observed at the same time (as noted above).

Behavioral Tests The trend within reproductive toxicology is to move from simple determinations of developmental landmarks and reflexes to more sophisticated and sensitive behavioral tests. This process was accelerated by the Environmental Protection Agency (EPA) of the United States, which issued guidelines requiring a “developmental neurotoxicity” study of compounds that meet any of several broad criteria. The behavioral tests to be performed in this study are extensive and rigidly defined. As laboratories become equipped and trained to meet these guidelines, they are adding such tests to their evaluations of pharmaceuticals. The suggestions for routine testing made below are considered reasonable for pre- and postnatal development studies intended as routine screens. It is suggested that testing be conducted on one or two adults per sex per litter, keeping the range of actual ages as tight as possible. Measurement of motor activity is commonly performed in the dark in cages or plastic boxes (open field) or residential mazes in which movement is quantitated by infrared detectors or by recording the interruption of light beams as the test subject moves through a horizontal grid of light beams. Possible parameters to evaluate include horizontal activity (light beams interrupted), number of movements, and time spent in the middle of the cage. The test period is selected to be long enough (normally 30–50 min) to allow the activity of the animals to decrease to an approximately constant level (asymptote). Testing of young pups (e.g., 13 days of age) is not recommended as their activity level is fairly constant during the test period and young unweaned pups should not be separated from their mothers for extended periods of time. Another test paradigm for detecting treatment effects on brain functioning in F1 offspring measures auditory startle habituation. In this test, the animal is placed in a chamber with a floor that detects movement. The animal is exposed to a sequence of 50 to 60 auditory stimuli, each at 110–120 decibels for 20–50 msec and separated by 5–20 sec. The gradual diminution of the animal’s movement response is indicative of normal habituation. There is not a consensus about the procedures to use to test for effects on learning and memory. The two most commonly used techniques are the water-filled maze, which

191

is preferred for measuring learning and passive avoidance, which is preferred for measuring memory (see Buelke-Sam et al., 1985). Retention is tested in a repeat test conducted approximately 1 week later.

Detecting Effects on Male Reproduction Male fertility studies with typical group sizes (15–30 males per group) are very insensitive for detecting effects on male fertility. If the control fertility rate is 80%, even a group size of 30 will only detect (at the 5% significance level) a 38% decrease in fertility 80% of the time and a 50% decrease 95% of the time (J. Antonello, personal communication, 1990). To detect slight effects on male fertility would require enormous group sizes. Mating each male with more than one female provides a more precise estimate of the reproductive capacity of each male but does not greatly increase statistical power. If multiple matings are to be done, it is recommended that the cohabitations with multiple females be sequential rather than concurrent. Not only is it difficult to detect effects on male fertility because of group-size considerations, effects on male fertility mediated by decreased sperm production are also difficult to detect because of the normally huge excess of sperm included in a rat ejaculate. Sperm production can be decreased by up to 90% without effect on fertility (either pregnancy rate or litter size) in the rat. This is not the case for men, so the sperm excess in the rat represents a serious flaw in the rat model (Working, 1988). To address this deficiency and improve the sensitivity of the model, it is advisable to determine the effects of the test agent on testes weights, testicular spermatid counts, and histopathology of the testes (preferably plastic sections) in the male fertility study and/or the 14-week toxicity study. In some cases, these parameters may be more predictive of possible effects on male fertility in humans than the fertility rate in rats.

DATA INTERPRETATION Use of Statistical Analyses Statistical analysis is a very useful tool for evaluating the effects of treatment on many developmental and reproductive toxicity parameters. For some parameters, such as maternal body weight changes, fetal weight, and horizontal activity in an open field, the comparison to the concurrent control is the primary consideration and, assuming adequate group size, the investigator relies heavily on the results of appropriate statistical analyses to interpret differences from control. For other parameters, though, statistical analysis is just one of several considerations that include historical control data and other relevant information about the test agent and related test agents. For example, statistical analysis of a low incidence of an uncommon fetal malformation will usually not be significant (p > 0.05) even if treatment related, due to the low power for detecting such effects with typical group sizes. In such cases, examination of the historical control data becomes paramount. If two fetuses with a particular malformation occur in separate litters only in a high-dose group, the finding is of much more concern if it is a very rare malformation than if recent historical control groups have had a few fetuses with that malformation. Other known effects of the test agent or related agents also sometimes contribute to data interpretation. For example, a low incidence of a malformation may be considered treatment related if it is at the low end of a typical

192

Safety Evaluation of Medical Devices

dose–response curve or if it is in a high-dose group and that malformation is an expected effect of the test agent. In general, though, a single occurrence of a fetal malformation in a treatment group (with none in control) is not cause for alarm, since this occurs in almost every study (together with occurrences of some malformations only in the control group). Statistical methods exist to appropriately analyze most developmental and reproductive toxicity parameters. Exceptions to this are the “r/m” litter parameters in which, for each litter, there is a number affected divided by the number in the litter. These parameters include preimplantation loss (r = corpora lutea-implants, m = corpora lutea), resorption rate (r = resorptions, m = implants), and the family of alteration rates (r = affected fetuses, m = fetuses). There are two factors complicating the statistical analysis of these data that have heretofore been inadequately handled (Clark et al., 1989). One is that almost all of these parameters have a strong dependence on m. For example, both preimplantation loss and resorption rate are normally higher at both the low and high extremes of m. In contrast, supernumerary rib tends to occur at higher incidences in average-size litters. The second factor that complicates the statistical analysis of r/m data is that affected implants tend to occur in clusters within litters (“litter effects”); that is, the intralitter correlation is greater than the interlitter correlation. For example, the total number of litters affected with anasarca, missing vertebra, and supernumerary rib is much less than would be expected by chance based on the number of affected fetuses. These problems have recently been resolved for analysis of resorption rate (and preimplantation loss) in SpragueDawley rats using a three-step process (Soper and Clark, 1991). First, based on an analysis of data from 1379 control rat litters examined since 1978, a likelihood score was derived for each (r, m) couplet based on the incidence of that couplet given that value of m. These scores were approximately equal to r. Second, an analysis of 136 litters from groups with slight effects on resorption rate revealed that, at low-effect doses of embryocidal test agents, the increases in resorptions tended to occur as increased proportion of affected litters. To maximize the difference in scores between control and affected litters, the scores for control-like litters (r = 1, 2, or 3) were down-graded from r(1, 2, and 3) to 0.4, 1, and 2.4, respectively. Third, to arrive at the final score for each litter, the modified r score for each litter was divided by the expected control value for that value of m. This last step makes the litter score immune to spontaneous or treatmentrelated effects on m. The final “robust” scores have more power for detecting effects than various other measures (raw r/m, affected litters/litters, r, Σr/Σn, and the likelihood score) and has a lower false positive rate with fluctuations in m. Covariance analysis (Snedecor and Cochran, 1967) can be used to reduce variability in a parameter and thereby increase sensitivity. For example, much of the variability in fetal weight data is due to variable litter size and, for rats, litters being sacrificed at different times during the workday. The variability due to these sources can be reduced by using litter size and time of sacrifice as potential covariates. Similarly, litter size and length of gestation can be used as covariates for neonatal pup weights and body weight at the beginning of treatment can be used as a covariate for maternal body weight changes during the treatment period of an embryo–fetal development study. A useful technique for determining if there is an effect of treatment on any toxicological parameter is the NOSTASOT method (Tukey et al., 1985; Antonello et al., 1993). This test is based on the principle that a possible toxicological effect of

interest occurs with a normal dose response; that is, there is an increasing effect with increasing dosage. The data to be analyzed should be examined first to confirm that this principle is not violated. In this method, regression analysis is used to determine if there is an increased or decreased response in a parameter with increasing dosage. This method can be visualized as a plot of response versus dosage in which the analysis determines if the slope of the plotted line deviates significantly from zero. This method can be used for essentially all parameters. Three analyses are performed—each with different spacing between dosage levels. The spacing in the first analysis is based on the arithmetic values of the dosage levels. The spacing in the second, referred to as the ordinal scaling, has equal spacing between dosage levels; that is, the control through high-dosage levels are assigned values of 0, 1, 2, and 3. In the third analysis, the log of the dosage level is used. Since the log of zero is impractical, the control group is assigned a value based on the spacing between the low- and middle-dosage levels according to a formula that assigns a log scale value to the control, such that the ratio of the difference between the control and low-dose groups and the difference between the low- and middle-dose groups is equal both in absolute values and in log scale values. This places the control group at a reasonable distance from the low-dosage group. The lowest p value among the three analyses—arithmetic, ordinal, and logarithmic—is taken as the p value of the overall analysis based on the assumption that, if there is a dosage-related effect, the method of analysis yielding the lowest value is the best model for that dosage response. A correction for the multiplicity of analyses can be applied. If none of the three analyses are significant at the 0.5 level, the analysis is complete and the high-dosage level is referred to as the “no statistical significance of trend dose,” or the NOSTASOT dose. If there is a significant trend through the high-dosage level, the data from the high-dosage level is deleted and the trend test repeated. This process is repeated until a NOSTASOT dose is determined. Effects at dosage levels above the NOSTASOT dose are then considered to be statistically significant. There are two major benefits of the NOSTASOT method. One is that spurious statistically significant results only at the low- and/or middle-dosage levels are eliminated, resulting in a reduction in false positives. A second benefit is that in some cases there may be real effects at multiple dosage levels that at any single dosage level are not statistically significant but will nevertheless result in a significant trend, thus providing increased sensitivity and halving the false negative rate.

Associations Between Developmental and Maternal Toxicity The developmental toxicity of many pharmaceuticals occurs only at maternally toxic dosages (Khera, 1984, 1985; Schardein, 1987). Also, there are several compounds for which there is evidence that their developmental toxicity is secondary to their maternal toxicity. The decreased uterine blood flow associated with hydroxyurea treatment of pregnant rabbits may account for the embryotoxicity observed. The teratogenicity of diphenylhydantoin in mice may be secondary to decreased maternal heart rate (Watkinson and Millicovsky, 1983) as supported by the amelioration of the teratogenicity by hyperoxia and the dependence on maternal genotype in genetic crosses between sensitive and resistant strains (Johnson et al., 1979). The hemolytic anemia caused in pregnant rabbits by diflunisal was severe enough to explain the concomitant axial skeletal malformations (Clark et al., 1984). Acetazolamide-induced fetal malformations in

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

mice are apparently related to maternal hypercapnia (Weaver and Scott, 1984a, b) and hypokalemia (Ellison and Maren, 1972). The increased resorption rate induced in rabbits by the antibiotic norfloxacin depends on exposure of the maternal gastrointestinal tract (Clark et al., 1986). In addition, various treatments that simulate effects that can result from pharmaceutical treatment have been shown to cause developmental toxicity. Food deprivation can cause embryo–fetal toxicity and teratogenicity in mice (Szabo and Brent, 1975; Hemm et al., 1977) and rats (Ellington, 1980) and fetal death, decreased fetal weight, and abortions in rabbits (Matsuzawa et al., 1981; Clark et al., 1986). Treatments that result in maternal hypoxia, such as hypobaric exposure (Degenhardt and Kladetzky, 1955) and blood loss (Grote, 1969), have been shown to be teratogenic. Also, the results from testing with numerous agents suggest that supernumerary rib in mice is caused by maternal stress (Kavlock et al., 1985; Beyer and Chernoff, 1986). Thus, in any case where developmental toxicity occurs at dosage levels with only moderate to severe maternal toxicity, the possibility of the developmental toxicity being secondary to the maternal toxicity can be considered. That is not to say, however, that it can be concluded that the developmental toxicity is secondary any time there is coincident maternal toxicity. To the contrary, it is usually very difficult to establish a causal relationship. Superficially similar types of maternal toxicity do not always cause the same pattern of developmental toxicity (see Chernoff et al., 1990). This may be because the developmental toxicity is secondary to maternotoxicity, but, since typical developmental toxicity studies include only a very cursory evaluation of maternal toxicity, the developmental toxicity may be secondary to an aspect of maternotoxicity that is not even being measured. To demonstrate that a developmental effect is secondary to a particular parameter of maternal toxicity, it is necessary but not sufficient to show that all mothers with developmental toxicity also had maternal toxicity and that the severity of the developmental effect was correlated with the maternal effect. Other examples where this approach has Table 13.6

193

been used to evaluate the relationship between maternal and developmental toxicity include: (1) the negative correlation between resorption rate and maternal body weight change in norfloxacin-treated rabbits (Clark et al., 1986), supporting the contention that the developmental toxicity was secondary; and (2) the lack of correlation between the embryotoxicity and maternal body weight change in pregnant mice treated with caffeine and L-phenylisopropyladenosine (Clark et al., 1987), suggesting no causal relationship. In many cases, additional studies specifically designed to address the relationship between developmental and maternal toxicity may be required.

In Vitro Alternatives The area of developmental toxicology actually is one of the earliest to have alternative models suggested for it, and has one of the most extensive and oldest literatures. This is, of course, partly owing to such models originally being used to elucidate the essential mechanisms and process of embryogenesis. Because of the complicated and multiphasic nature of the developmental process, it has not been proposed that any of these systems be definitive tests, but rather that they serve as one form or another of a screen. As such, these test systems would either preclude or facilitate more effective full-scale evaluation in one or more of the traditional wholeanimal test protocols. The literature and field are much too extensive to review comprehensively here. There are a number of extensive review articles and books on the subject (Wilson, 1978; Clayton, 1981; Kochhar, 1981; Saxen, 1984; Homburger and Goldberg, 1985; Faustman, 1988; Daston and D’Amato, 1989), which should be consulted by those with in-depth interest. The existing alternative test systems fall into six broad classes: (1) lower organisms; (2) cell culture systems; (3) organ culture systems; (4) submammalian embryos; (5) mammalian embryos; (6) others. Table 13.6 provides an overview of the major representatives of these six groups, along with at least one basic

Alternative Developmental Toxicity Test Systems

Category

Test system

Model

References

I: Lower organisms:

Sea urchins Drosophila Trout Planaria Brine shrimp

Organism Intact and embryonic cells (Fish species) Regeneration Disruption of elongation: DNA and protein levels in Artemia nauplii

Kotzin and Baker (1972) Abrahamson and Lewis (1971) MacCrimmon and Kwain (1969) Best et al. (1981) Kerster and Schaeffer (1983); Sleet and Brendel (1985)

II: Cell culture

Protein synthesis of cultured cells Avian neural crest Neuroblastoma Lectin-mediated attachment

Pregnant mouse and chickens’ epithelial cells Differentiation of cells Differentiation of cells Tumor cells

Clayton (1979) Sieber-Blum (1985) Mummery et al. (1984) Braun and Horowicz (1983)

III: Organ culture

Frog limb Mouse embryo limb bud

Bazzoli et al. (1977) Kochhar and Aydelotte (1974)

Metanephric kidney organ cultures

Regeneration Inhibition of incorporation of precursor and of DNA synthesis From 11 day mouse embryos

IV: Submammalian embryo

Chick embryo Frog embryo

Xenopus laevis

Rat embryo culture

Whole postimplantation embryos

Chernoff ‘Micromass cultures’

Mouse embryo short test Rat embryo midbrain and limb

Brown and Fabro (1981); Cockroft and Stele (1989) Chernoff and Kavlock (1980) Flint and Orton (1984)

Structure-activity relationships (SAR)

Mathematical correlations of activity with structural features

Einslein et al. (1983b) Gombar et al. (1990)

V: Mammalian embryo

VI: Other

Saxen and Saksela (1971) Gebhardt (1972) Davis et al. 1981)

194

Safety Evaluation of Medical Devices

Table 13.7

Developmental Toxicity Test System Considerations

Possibility

in vivo

Organ culture

Cell structure

Lower organisms

Mammalian embryo culture

Submammalian embryos

Other

To study maternal organ factors To study embryogenesis as a whole To eliminate maternal confounding factors (nutrition, etc.) To eliminate placental factors (barrier differences) To study morphogenetic events To create controllable, reproducible conditions For exact exposure and timing For microsurgical manipulations For continuous registration of the effects To collect large amounts of tissue for analysis To use human embryonic tissue for testing Screening

Yes Yes No

No No Yes

No No Yes

No No No

No/Yes Yes Yes

No/Yes Somewhat Yes

NA NA NA

No

Yes

Yes

No

Yes

No

NA

Difficult Difficult Difficult Difficult Difficult Yes No Expensive

Yes Yes Yes Yes Yes Difficult Yes Yes

No Yes Yes No Yes Yes Yes Yes

Maybe Yes Yes Maybe No No No Yes

Yes Yes Yes Yes Yes Yes No Yes

Yes Yes Yes Yes Yes No No Yes

NA NA NA NA NA NA NA Yes

NA = not available.

reference to the actual techniques involved and the system components for each. The comparative characteristics of these different classes of test systems are presented in Table 13.6. The key point is that these systems can be used for a wide range of purposes, only one of which is to screen compounds to determine the degree of concern for developmental toxicity. The utility of these systems for screening is limited by the degree of dependability in predicting effects primarily in people and secondarily in the traditional whole-animal test systems. Determining the predictive performance of alternative test systems requires the evaluation of a number of compounds for which the “true” (human) effect is known. In 1983 a consensus workshop generated a so-called “gold standard” set of compounds of known activity (Smith et al., 1989). The composition of this list has been open to a fair degree of controversy over the years (Flint, 1989; Johnson, 1989; Johnson et al., 1989). However, an agreed-upon “gold standard” set of compounds of known activity is an essential starting point for the validation of any single test system or battery of test systems because of the multitude of mechanisms for developmental toxicity. It is unlikely that any one system will be able to stand in place of segment II studies in two species, much less to accurately predict activity in humans. Their use as general screens or as test systems with little potential for extensive or intended human exposure will, however, probably be appropriate. Table 13.7 presents the relative merits and utility of these test systems.

REFERENCES Abrahamson, S. and Lewis, E.B. (1971). The detection of mutations in Drosophila melanogaster. In: A. Hollanender (Ed.), Chemical Mutagens. Principles and Methods of Their Detection, Vol. 2. Plenum Press, New York, pp. 461–488. Anderson, C. and Clark, R.L. (1990). External genitalia of the rat: Normal development and the histogenesis of 5α-reductase inhibitor-induced abnormalities. Teratology, 42: 483–496. Antonello, J.M., Clark, R.L. and Heyse, J.F. (1993). Application of the Tukey trend test procedure to assess developmental and reproductive toxicity. I. Measurement data. Fund. Appl. Tox., 21: 52–58. APHIS (1989). Animal and Plant Health Inspection Service, V. S Department of Agriculture, Fed. Reg., 54(168): 36112–36163.

Arnold, D. L., Grice, H. C. and Krewski, D. R. (1990) Handbook of In Vivo Toxicity Testing, Academic Pren, Inc., San Diego. Bazzoli, A.S., Manson, J., Scott, W.J. and Wilson, J.B. (1977). The effects of thalidomide and two analogues on the regenerating forelimb of the newt. J. Embryol. Exptl Morphol., 41: 125–135. Best, J.B., Moritia, M., Ragin, J., and Best, J., Jr. (1981). Acute toxic responses of the freshwater planarian, Dugesia dorothocephala, to methylmercury. Bull. Environ. Contam. Toxicol., 27: 49–54. Beyer, P.E. and Chernoff, N. (1986). The induction of supernumeraryribs in rodents: the role of maternal stress. Teretog. Carcinog. Utag., 6: 419–429. Bloch, E., Lew, M. and Klein, M. (1971). Studies on the inhibition of fetal androgen formation. Inhibition of testosterone synthesis in rat and rabbit fetal testes with observations on reproductive tract development. Endocrinology, 89: 16–31. Braun, A.G. and Horowicz, P.B. (1983). Lectin-mediated attachment assay for teratogens. Results with 32 pesticides. J. Toxicol. Environ. Hlth., 11(2): 275–286. Brown, N.A. and Fabro, S. (1981). Quantitation of rat embryonic development in vitro: A morphological scoring system. Teratology, 24: 65–78. Buelke-Sam, J., Kimmel, C.A. and Adams, J. (1985). Design considerations in screening for behavioral teratogens: Results of the collaborative behavioral teratology study. Neurobehavioral Toxicology and Teratology, 7: 537–789. Chapin, R.E., Gulati, D.R. and Barnes, L.H. (1991). The effects of dietary restriction on reproductive endpoints in Sprague-Dawley rats. Toxicologist, 11: 112. Chernoff, N. and Kavlock, R.J. (1980). A potential in vivo screen for the determination of teratogenic effects in mammals. Teratology, 21: 33A–34A. Chernoff, N., Setzer, R.W., Miller, D.B., Rosen, M.B. and Rogers, J.M. (1990). Effects of chemically induced maternal toxicity on prenatal development in the rat. Teratology, 42: 651–658. Chow, B.F. and Rider, A.A. (1973). Implication of the effects of maternal diets in various species. J. Anim. Sci., 36: 167–173. Christian, M.S. (1983). Assessment of reproductive toxicity: State of the art. In: M.S. Christian, M. Galbraith, P. Voytek and M.A. Mehlman (Eds.), Assessment of Reproductive and Teratogenic Hazards. Princeton Scientific Publishers, Princeton, pp. 65–76. Christian, M.S. and Hoberman, A.M. (1989). Current in vivo reproductive toxicity and developmental toxicity (teratology) test methods. In: J.A. Marquis and A.W. Hayes (eds.), A Guide to General Toxicology, Second Edn. S. Karger, Basel, Switzerland, pp. 91–100. Clark, R.L., Robertson, R.T., Minsker, D.H., Cohen, S.M., Toco, D.J., Allen, H.L., James, M.L. and Bokelman, D.L. (1984). Diflunisalinduced maternal anemia as a cause of teratogenicity in rabbits. Teratology, 30: 319–332.

Chapter 13:

Subchronic and Chronic Toxicity and Reproductive and Developmental Toxicity

Clark, R.L., Robertson, R.T., Peter, C.P., Bland, J.A., Nolan, T.E., Oppenheimer, L. and Bokelman, D.L. (1986). Association between adverse maternal and embryo-fetal effects in norfloxacin-treated and food-deprived rabbits. Fund. Appl. Toxicol., 7: 272–286. Clark, R.L., Eschbach, K., Cusick, W.A. and Heyse, J.F. (1987). Interactions between caffeine and adenosine agonists in producing embryo resorptions and malformation in mice. Toxicol Appl Pharmacol., 91(3): 371–385. Clark, R.L., Antonello, J.M., Soper, K.A., Bradstreet, T.E., Heyse, J.F., and Ciminera, J.L. (1989). Statistical analysis of developmental toxicity data. Teratology, 39: 445–446. Clark, R.L., Anderson, C.A., Prahalada, S., Leonard, Y.M., Stevens, J.L. and Hoberman, A.M. (1990a). 5α-Reductase inhibitor-induced congenital abnormalities in male rat external genitalia. Teratology, 41: 544. Clark, R.L., Antonello, J.M., Grossman, J.T., Wise, L.D., Anderson, C., Bagdon, W.J., Prahalada, S., MacDonald, J.S. and Robertson, R.T. (1990b). External genitalia abnormalities in male rats exposed in utero to finasteride, a 5α-reductase inhibitor. Teratology, 42: 91–100. Clayton, R.M. (1979). In: A.N. Rowan, and C.J. Stratmann, (Eds.), Alternatives in Drug Research. Macmillan Press, London, p. 153. Clayton, R.M. (1981). An in vitro system for teratogenicity testing. In: A.N. Rowan, and C.J. Stratmann, (Eds.), The Use of Alternatives in Drug Research. University Park Press, Baltimore, pp. 153–173. Cockroft, D.L. and Steele, C.E. (1989) Postimplantation embryo culture and its application to problems in teratology. In: C.K. Atterwill, and C.E. Steele, (Eds.), In Vitro Methods in Toxicology. Cambridge University Press, New York, pp. 365–389. Daston, D.L. and D’Amato, R.A. (1989). In vitro techniques in teratology. In: M. Mehlman, (Ed.), Benchmarks: Alternative Methods in Toxicology. Princeton Scientific, Princeton, New Jersey, pp. 79–109. Davis, K.R., Schultz, T.W. and Dumont, J.N. (1981). Toxic and teratogenic effects of selected aramatic amines on embryos of the amphibian Xenopus laevis. Arch. Environ. Contam. Toxicol., 10: 371–391. Degenhardt, K. and Kladetzky, J. (1955). Spinal deformation and chordal attachment. Z. Menschl. Vererb. Konstitutionsl., 33: 151–192. Ellington, S. (1980). In vivo and in vitro studies on the effects of maternal fasting during embryonic organogenesis in the rat. J. Reprod. Fertil., 60: 383–388. Ellison, A.C. and Maren, T.H. (1972). The effect of potassium metabolism on acetazolamide-induced teratogenesis. Johns Hopkins Med J., 130(2): 105–15. Enslein, K., Lander, T.R. and Strange, J.L. (1983b). Teratogenesis: A statistical structure-activity model. Terat. Carcin Mutagen, 3: 289–309. Faustman, E.M. (1988). Short-term tests for teratogogens. Mutation Res., 205: 355–384. Flint, O.P. (1989). Reply to Letter to the Editor. Toxicol. Appl. Pharmacol., 99: 176–180. Flint, O.P. and Orton, T.C. (1984). An in vitro assay for teratogens with cultures of rat embryo midbrain and limb bud cells. Toxicol. Appl. Pharmacol., 76: 383–395. Food and Drug Administration (1983). Good Laboratory Practices for Nonclinical Laboratory Studies. CFR 21 Part 58, March, 1983. Food and Drug Administration (1988). Good Laboratory Practices, CFR 21 Part 58, April, 1988. Gad, S.C. (2006). Animal Models in Toxicology, 2nd Edn. Marcel Dekker, New York. Gad, S.C. and Chengelis, C.P. (1999). Acute Toxicology, 2nd Ed. Academic Press, San Diego, CA. Gebhardt, D.O.E. (1972). The use of the chick embryo in applied teratology. In D.H.M. Woollam (Ed.), Advances in Teratology, Vol. 5. Academic Press, London, pp. 97–111. Gombar, V.K., Borgstedt, H.H., Enslein, K., Hart, J.B. and Blake, B.W. (1990). A QSAR model of teratogenesis. Quant. Struct.–Activ. Rel., 10: 306–332. Grote, W. (1969). Trunk skeletal malformations following blood loss in gravid rabbits. Z. Anat. Entwicklungsgesch., 128: 66–74.

195

Gulati, W. (1969). Trunk skeletal malformations following blood loss in gravid rabbits. Z. Anat. Entwicklungsgesch., 128: 66–74 Habert, R. and Picon, R. (1984). Testosterone, dihydrostestosterone and estadiol-17β levels in maternal and fetal plasma and in fetal testes in the rat. J. Steroid Biochem., 21: 183–198. Haschek, W.M. and Rousseaup, C.G. (1991). Handbook of Toxicology Pathology. Academic Press, San Diego. Heinrichs, W.L. (1985). Current laboratory approaches for assessing female reproductive toxicity. In: R.L. Dixon (Ed.), Reproductive Toxicology. Raven Press, New York, pp. 95–108. Hemm, R., Arslanoglou, L., and Pollock, J. (1977). Cleft palate following prenatal food restriction in mice: Association with elevated maternal corticosteroids. Teratology, 15: 243–248. Heywood, R. and James, R.W. (1985). Current laboratory approaches for assessing male reproductive toxicity: Testicular toxicity in laboratory animals. In: R.L. Dixon (Ed.), Reproductive Toxicology, Report. Environ. Health Perspect, pp 147–160. Homburger, J. and Goldberg, A.M. (1985). In Vitro Embryotoxicity and Teratologenicity Tests. S. Karger AG, Basel, Switzerland. Hood, R.D. (Ed.) (2007) Developmental and Reproductive Toxicology, 2nd Edn. Taylor & Francis, New York. Johnson, E.M. (1989). Problems in validation of in vitro developmental toxicity assays. Fund. Appl. Toxicol., 13: 863–867. Johnson, E.M., Gorman, R.M., Gabel, B.E.C. and George, M.E. (1982). The Hydra attenuata system for detection of teratogenic hazards. Terat. Carcin. Mutagen, 2: 263–276. Johnson, E.M., Newman, L.M. and Fu, L. (1989) Letter to the Editor. Toxicol. Appl. Pharmacol., 99: 173–176. Johnston, M.C., Sulik, K.K. and Dudley, K.H. (1979). Genetic and metabolic studies of the differential sensitivity of A/J and C57BL/6J mice to phenytoin (‘Dilantin’)-induced cleft lip. Teratology, 19: 33A. Kavlock, R.J., Chernoff, N. and Rogers, E.H. (1985). The effect of acute maternal toxicity on fetal development in the mouse. Teratog. Carcinog. Mutag., 5: 3–13. Kerster, H.W. and Schaeffer, D.J. (1983). Brine shrimp (Artemia salina) Nauplia as a teratogen test system. Ecotoxicol. Environ. Safety, 7: 342–349. Khera, K.S. (1984). Maternal toxicity—A possible factor in fetal malformations in mice. Teratology, 29: 411–416. Khera, K.S. (1985). Maternal toxicity: A possible etiological factor in embryo-fetal deaths and fetal malformations of rodent-rabbit species. Teratology, 31: 129–153. Kochhar, D.M. (1981). Embryo explants and organ cultures in screening of chemicals for teratogenic effects. In: C.A. Kimmel, and J. Buelbe-Saw, (Eds.), Developmental Toxicology. Raven Press, New York, pp. 303–319. Kochhar, D.M. and Aydelotte, M.B. (1974). Susceptible stages and abnormal morphogenesis in the developing mouse limb, analyzed in organ culture after transplacental exposure to vitamin A (retinoic acid). J. Embryol. Exptl. Morphol., 31: 721–734. Korenbrot, C.C., Huhtaniemi, I.T. and Weiner, R.I. (1977). Preputial separation as an external sign of pubertal development in the male rat. Biol. Reprod., 17: 298–303. Kotzin, B.L. and Baker, R.F. (1972). Selective inhibition of genetic transcription in sea urchin embryos. J. Cell. Biol., 55: 74–81. Lochry, E.A., Hoberman, A.M. and Christian, M.S. (1984). Positive correlation of pup body weight with other commonly used developmental landmarks. Teratology, 29: 44A. Lochry, E.A., (1987). Concurrent use of behavioral/functional testing in reproductive and developmental toxicity screen: Practical considerations. J. Am. Coll. Toxicol., 6: 433–9. MacCrimmon, H.R. and Kwain, W.H. (1969). Influences of light on early development and meristic characters in the rainbow trout (Salmo gairdneri Richardson). Can. J. Zool., 47: 631–637. Manson, J.M. and Kang, Y.J. (1989). Test methods for assessing female reproductive and developmental toxicology. In: A.W. Hayes (Ed.), Principles and Methods of Toxicology, 2nd Edn. Raven Press, New York, pp. 311–359. Matt, D.W., Lee, J., Sarver, P.L., Judd, H.L. and Lu, J.K.H. (1986). Chronological changes in fertility, fecundity, and steroid hormone

196

Safety Evaluation of Medical Devices

secretion during consecutive pregnancies in aging rats. Biol. Reprod., 34: 478–487. Matt, D.M., Sarver, P.L. and Lu, J.K.H. (1987). Relation of parity and estrous cyclicity to the biology of pregnancy in aging female rats. Biol. Reprod., 37: 421–430. Matsuzawa, T., Nakata, M., Goto, I. and Tsushima, M. (1981). Dietary deprivation induces fetal loss and abortions in rabbits. Toxicology, 22: 255–259. Mulay, S., Varma, D.R. and Soloman, S. (1982). Influence of protein deficiency in rates on hormonal status and cytoplasmic glucocorticoid receptors in maternal and fetal tissues. J. Endocrin., 95: 49–58. Mummery, C.L., van den Brink, C.E., van der Saag, P.T. and de Loat, S.W. (1984). A short-term screening test for teratogens using differentiating neuroblastoma cells in vitro. Teratology, 29: 271–279. Palmer, A.K. (1981). Regulatory requirements for reproductive toxicology: Theory and practice. In: C.A. Kimmel and J. BuelkeSam (Eds.), Developmental Toxicology. Raven Press, New York, pp. 259–287. Persaud, T.V.N. (1985). Teratogenicity testing. In: T.V.N. Persaud, A.E. Chudley and R.G. Skalko (Eds.), Basic Concepts in Teratology. Alan R. Liss, New York, pp. 155–181. Saegusa, T., Kaneko, Y., Sato, T., Nrama, I. and Segima, Y. (1980). BD40A-induced wavy ribs in rats. Teratology, 22: 14A. Salewski, V.E. (1964). Färbethode zum makroskopischen Nachweis von Implantationsstellen am Uterus der Ratte. NauynSchmiedebergs. Arch. Exp. Path. U. Pharmak., 247: 367. Saxen, L. and Saksela, E. (1971). Transmission and spread of embryonic induction. II. Exclusion of an assimilatory transmission mechanism in kidney tubule induction. Exptl. Cell Res., 66: 369–377. Schardein, J. (1987). Approaches to defining the relationship of maternal and developmental toxicity. Teratog. Carcinog. Mutag., 7: 255–271. Schardein, J. (1988). Teratologic testing: Status and issues after two decades of evolution. Rev. Environ. Contam. Toxicol., 102: 1–78. Sieber-Blum, M.F. (1985). Differentiation of avian neural crest cells in vitro (quail, chick, rodent). Crisp Data Base, HD15311-04. Sleet. R.B. and Brendel, K. (1985). Homogenous populations of Artemia nauplii and their potential use for in vitro testing in developmental toxicology. Terat. Carcin. Mutagen, 5(1): 41–54. Smith, M.K., Kimmel, G.L., Kochhar, D.M., Shepard, T.H., Spielberg, S.P. and Wilson, J.C. (1989). A selection of candidate compounds for n vitro teratogenesis test validation. Terat. Carcin. Mutagen., 3: 461–480. Snedecor, G.W. and Cochran, W.G. (1967). Statistical Methods, 6th Edn, chapter 10. Iowa State University Press, Ames. Sopor, K.A. and Clark, R.L. (1990). Exact permutation trend tests for fetal survival data. Proc. Biopharm. Section Am. Stat. Assoc., pp. 263–268. Staples, R.E. (1971). Blastocyst transplantation in the rabbit. In: J.C. Daniel, Jr., (Ed.), Methods in Mammalian Embryology. W.H. Freeman, San Francisco, pp. 290–304.

Szabo, K. and Brent, R. (1975). Reduction of drug-induced cleft palate in mice. Lancet (June), 1296–1297. Tanimura, T. (1990). The Japanese perspectives on the reproductive and developmental toxicity evaluation of pharmaceuticals. J. Am. Coll. Toxicol., 9: 27–38. Tukey, J.W., Ciminera, J.L. and Heyse, J.F. (1985). Testing the statistical certainty of a response to increasing doses of a drug. Biometrics, 41: 295–301. Turner, J.R. (1973). Perinatal mortality, growth, and survival to weaning in offspring of rats reared on diets moderately deficient in protein. Br. J. Nutr., 29: 139–147. Tyl, R.W. (1988). Developmental toxicity in toxicological research and testing. In: B. Ballantyne (Ed.), Perspectives in Basic and Applied Toxicology. Butterworth, pp. 206–241. Watkinson, W.P., and Millocovsky, G. (1983). Effect of phenytoin on maternal heart rate in A/J mice: Possible role in teratogenesis. Teratology, 28: 1–8. Weaver, T.E. and Scott, W.J. (1984a). Acetazolamide teratogenesis: Association of maternal respiratory acidosis and ectrodactyly in C57BL/6J mice. Teratology, 30: 187–193. Weaver, T.E., and Scott, J.J. (1984b). Acetazolamide teratogenesis: Interaction of maternal metabolic and respiratory acidosis in the induction of ectrodactyly in C57BL/6J mice. Teratology, 30: 195–202. Weindruch, R. and Walford, R.L. (1988). The Retardation of Aging and Disease by Dietary Restriction. Charles C. Thomas, Springfield, Illinois. Wickramaratne, G.A. de S. (1988). The post-natal fate of supernumerary ribs in rat teratogenicity studies. J. Appl. Toxicol., 8: 91–94. Wilson, J.G. (1978). Review of in vitro systems with potential for use in teratogenicity screening, J Environ Pathol Toxicol, 2(1): 149–167. Wilson, J. (1965). Methods for administering agents and detecting malformations in experimental animals. In: J.J. Wilson and J. Warkany (Eds.), Teratology, Principles and Techniques. The University of Chicago Press, Chicago, pp. 262–277. Wise, L.D., Clark, R.L., Minsker, D.H. and Robertson, R.T. (1988). Use of hematology and serum biochemistry data in developmental toxicity studies. Teratology, 37: 502–503. Wise, L.D., Clark. R.L., Rundell, J.O. and Robertson, R.T. (1990a). Examination of a rodent limbbud micromass assay as a prescreen for developmental toxicity. Teratology, 41: 341–351. Wise, L.D., Vetter, C.M., Anderson, C., Antonello, J.M. and Clark, R.L. (1990b). Reversible effects on external genitalia in rats exposed in vitro to triamcinolone acetonide. Teratology, 41: 600. Working, P.K. (1988). Male reproductive toxicology: Comparison of the human to animal models. Environ. Health. Perspec., 77: 37–44. Yamada, T., Ohsawa, K. and Ohno, H. (1988). The usefulness of alkaline solutions for clearing the uterus and staining implantation sites in rates. Exp. Anim. (Tokyo), 37: 325–332. Zeman, F.J. (1967). Effect on the young rat of maternal protein restriction. J. Nutr., 93: 167–173.

14 Carcinogenicity

This chapter studies the potential tumorigenicity and carcinogenicity of devices and biomaterials with prolonged human exposure via implantation. Carcinogenicity studies are infrequently required for medical devices. The tables in Chapter 1 cite each of the general cases though these do not catch all of the nuances. Under ISO 10993-11, for example, the need to perform carcinogenicity tests may be triggered by: ● ● ● ●

but this is now so infrequent we will concentrate on the case of the rat study in this section. The use of a single species is unlikely to adversely effect overall ability to detect potential risks (Zbinden, 1993). The choice of species and strain to be used in a carcinogenicity study is based on various criteria, including susceptibility to tumor induction, incidence of spontaneous tumors survival, existence of an adequate historical data base, and availability (Cameron et al., 1985; Arnold et al., 1990; Gad, 2002). Susceptibility to tumor induction is an important criterion. There would be little justification for doing carcinogenicity studies in an animal model that did not respond when treated with a “true” carcinogen. Ideally, the perfect species/strain would have the same susceptibility to tumor induction as the human. Unfortunately, this information is usually unavailable, and the tendency has been to choose animal models that are highly sensitive to tumor induction to minimize the probability of false negatives. The incidence of spontaneous tumors is also an important issue. Rodent species and strains differ greatly in the incidence of various types of spontaneous tumors. The Sprague-Dawley stock, although preferred, has a very high incidence of mammary tumors in aging females, which results in substantial morbidity during the second year of a carcinogenicity study. If one chooses the Fischer 344 (F344) strain, the female mammary tumor incidence will be lower, but the incidence of testicular tumors will be higher (close to 100%), than that in Sprague-Dawley rats. A high spontaneous tumor incidence can compromise the results of a carcinogenicity study in two ways. If a compound induces tumors at a site that already has a high spontaneous tumor incidence, it may be impossible to detect an increase above the high background “noise.” Conversely, if a significant increase above levels is demonstrated, one may question the relevance of this finding to humans on the basis that the species is “highly susceptible” to tumors of this type (Hajian, 1983). Such considerations are further compounded by the “Oppenheimer Effect” (Turner, 1941; Oppenheimer et al., 1948, 1952, 1953, 1955, 1958, 1961, 1964). This is the occurrence of parenchymal tumors produced after long lasting periods when smooth-surfaced solids are implanted. Such solids have included everything from plastics to marble chips. This is a well-established phenomenon in rodents which has not been demonstrated in nonrodents or humans. These tumors are thought to be due to a epigenetic mechanism, and no sex differences in response has been seen. About 80% of the resulting tumors are fibrosarcomas (Alexander and Horning, 1959; Brand et al., 1975, 1976; Ecanow et al., 1977; Brand and Brand, 1980; Memol, 1986). Particulate “generation” by the

Devices introduced in the body >30 days cumulative contact Devices or materials with positive genotoxicity tests Resorbable materials and devices Note that in those cases where carcinogenicity testing is required but no effects have occurred in genotoxicity tests, clinical testing may be performed concurrently with carcinogenicity testing. Where implantation does not represent the most appropriate route of exposure, scientifically justified alternatives should be considered.

Even then, the ISO standard states that “carcinogenicity should be conducted only if there are suggestive data from other sources.” Where implantation does not represent the most appropriate or there is a more practical route of exposure, scientifically justified alternative routes should be considered (Henry, 1985). The intent of such testing is to determine the carcinogenic (“tumorigenic”) potential of devices, materials and/or multiple exposures over a period of the total life-span of the test animal. Such tests are frequently designed to evaluate both the chronic toxicity and the tumorigenicity in a single study as well as device efficacy. These studies are the longest and most expensive of the nonclinical studies typically conducted on any new device or device material. These studies are important because, as noted by the International Agency for Research on Cancer (1987), “in the absence of adequate data on humans, it is biologically plausible and prudent to regard agents for which there is sufficient evidence of carcinogenicity in experimental animals as if they presented a carcinogenic risk to humans.” The best established risks of carcinogenicity have to do with the effects of metals leading from long term implants.

ANIMAL MODEL Unlike for pharmaceuticals and agrichemicals, only one species is required to be evaluated in a carcinogenicity assay for a device or device material. The Sprague-Dawley derived rat is by far the most commonly used with the other strains of rats (Wistar, Long Evans, CFE, and Fischer 344 seeing only rare use). On very rare occasions dogs have been used (besides other concerns, a dog tumorigenial study is required to run seven years, as opposed to two for a rat, to be valid), 197

198

Safety Evaluation of Medical Devices

degradation of device components increases the degree of problems. From these early investigations, one can derive a number of characteristics for the phenomenon termed “solid-state carcinogenesis.” The major ones are: (a) Composition of the material per se appears to be of little importance (unless it contains leachable carcinogens) because a wide variety of materials elicits a similar response. (b) A continuous, impermeable surface is important since perforations, weaves, or powders tend to reduce or abolish tumorigenicity of the material (Bates and Klein 1966; Bischoff and Bryson, 1964; Dukes and Mitchley, 1962; Goldhaber, 1961, 1962). (c) The implant must be of at least a minimum (“critical” size). (d) The implant must remain in situ for a minimum period of time. The studies of Oppenheimer et al. (1958) found the presarcomatous changes occurred when the material was in place for about six months, although tumors may not appear for many more months. The ability of a species/strain to survive for an adequate period is essential for a valid assessment of carcinogenicity. Poor survival has caused regulatory problems for pharmaceutical companies (PMA, 1988) and is, therefore, an important issue for medical devices. The underlying concept is that animals should be exposed to the drug for the greater part of their normal life span to make a valid assessment of carcinogenicity. If animals on study die from causes other than drug-induced tumors, they may not have been at risk long enough for tumors to have developed. The sensitivity of the bioassay would be reduced and the probability of a false negative result would be increased. The availability of an adequate historical data base if often cited as an important criterion for species/strain selection. Historical control data can sometimes be useful in evaluating the results of a study. Although such data are not considered equal in value to concurrent control data, they can be helpful if there is reason to believe that the concurrent control data are “atypical” for the species/strain. Advantages of the Sprague-Dawley rat are (1) a large historical data base including various routes of exposure, (2) demonstrated susceptibility to known carcinogens, (3) generally good survival until recently (see below), and (4) ease of handling compared with certain other stocks. Disadvantages include (1) moderate to high incidence of spontaneous tumors, especially mammary and pituitary, (2) old rat nephropathy, and (3) marked genetic variability in stocks obtained from different suppliers (Chu et al., 1981; Sher et al., 1982). There has recently been a reduction in survival of Sprague-Dawley rats and rats of other strains (Food and Drug Administration (FDA), 2001). This reduction may be the result of ad libitum feeding, as preliminary results suggest that caloric restriction may improve survival. Leukemia appears to be the major cause of decreasing survival in the F344 rat. The problem of reduced survival may necessitate a reevaluation of the survival requirements for carcinogenicity studies by regulatory agencies. There is also now a significant body of data that suggest that switching from the long favored ad libritum feeding of animals in bioassays can both extend their life-span and decrease the incidences of some background tumors (Rao and Huff, 1990).

DOSE SELECTION Number of Dose Levels Unlike for drugs or agricultural chemicals, there will ordinarily be two dose levels, the maximum implantable dose (MID), and a fraction thereof (usually one half of the MID). The controls will generally include polyethylene implants or other materials whose lack of carcinogenic potential is documented in a comparable form and shape. In carcinogenicity testing on rodents, the maximum implantable dose (MID) of a material or device should be applied. Where possible, this dose should be expressed as a multiple of the worst case human exposure in milligrams per kilogram.

Group Size The minimum number of animals assigned to each dose group in implant carcinogenicity studies is 50 of each sex. Most companies, however, use more than the minimum number, and some use up to 80 animals per sex per group. The most important factor in determining group size is the need to have an adequate number of animals for a valid assessment of carcinogenic activity at the end of the study. Larger group sizes are also used when the carcinogenicity study is combined with a chronic toxicity study in the rat. In this case, serial sacrifices are performed at 6 and 12 months to evaluate potential toxic effects of the device. In the final analysis, the sensitivity of the bioassay for detecting carcinogens is directly related to the sample size. Use of the MTD has often been justified based on the small number of animals at risk compared to the potential human population, in spite of the difficulties inherent in extrapolating effects at high doses to those expected at much lower clinical doses. A reasonable compromise may be the use of doses lower than the MTD combined with a larger group size than the 50 per sex minimum accepted by regulatory agencies.

Route of Administration Device carcinogenicity studies are conducted with the device or material being implanted into the test animals. Prior to implantation the samples are prepared. Whenever possible, the device shall be tested in its “ready-to-use” form. Otherwise a suitably formed implant shall be made of the test material, with appropriate consideration of potential solid state carcinogenicity. Treated animals typically receive single implants in a flank by making an incision, opening a pouch, inserting the sample, and closing the pouch. Dose groups are achieved by implanting variable numbers of devices in multiple flanks. Controls are generally untreated in the sense that no device is implanted—only the surgical procedure is performed.

Study Duration The duration of carcinogenicity studies for rats is two years. Occasionally, rat studies are extended to 30 months. When hamsters are used, the study duration is limited to 18 months, a time period that is consistent with the shorter survival characteristics of this species. Irrespective of the intended duration of the study, the most important consideration is that adequate numbers of animals survive long enough to allow for a valid assessment of carcinogenic activity. When survival is problematic, the duration of the study may be modified accordingly. The effect of survival on study duration is discussed in the next section.

Chapter 14:

Survival As stated earlier, adequate survival is of primary importance in carcinogenicity studies because animals must be exposed to a drug for the greater part of their life span to increase the probability that late-occurring tumors can be detected. Early mortality, resulting from causes other than tumors, can jeopardize the validity of a study because dead animals cannot get tumors. In general, the sensitivity of a carcinogenicity bioassay is increased when animals survive to the end of their natural life span, because weak carcinogens may induce late-occurring tumors. The potency of a carcinogen is often inversely related to the time to tumor development. By analogy, as the dose of a carcinogen is reduced, the time to tumor occurrence is increased (Littlefield et al., 1979; DePass et al., 1986). Why do we not allow all animals on a carcinogenicity study to live until they die a natural death if by so doing we could identify more drugs as carcinogens? In fact, the sensitivity of a bioassay may not be improved by allowing the animals to live out their natural life span because the incidence of spontaneous tumors tends to increase with age. Thus, depending on the tumor type, the ability of the bioassay to detect a device-related increase in tumor incidence may actually decrease, rather than increase, with time. Therefore, the optimum duration of a carcinogenicity study is that which allows late-occurring tumors to be detected but does not allow the incidence of spontaneous tumors to become excessive. Reduced survival in a carcinogenicity study may or may not be device-related. Sometimes, the MTD is exceeded and increased mortality occurs at the highest dose level and, occasionally, at the mid-dose level as well. This situation may not necessarily invalidate a study; in fact, the protocol may be amended to minimize the impact of the device-induced mortality. For example, cessation of drug treatment may enhance the survival of the animals in the affected groups, and allow previously initiated tumors to develop. As shown by Littlefield et al. (1979) in the NCTR ED01 study, liver tumors induced by 2-acetylaminofluorene, which appeared very late in the study, were shown to have been induced much earlier and not to require the continuous presence of the carcinogen to develop. By contrast, bladder tumors that occurred in the same study were dependent on the continued presence of the carcinogen.

Table 14.1 ● ●

● ●

●

Carcinogenicity

199

Whether drug treatment is terminated or not, devicerelated toxicity may also be managed by performing complete histopathology on animals in the lower-dose groups rather than on high-dose and control animals only. If there is no increase in tumor incidence at a lower-dose level that is not compromised by reduced survival, the study may still be considered valid as an assessment of carcinogenicity. When reduced survival is related to factors other than excessive toxicity, the number of animals at risk for tumor development may be inadequate, and the validity of the study may be compromised even in the absence of a device effect on survival. Obviously, the adjustments described above for excessive, drug-related toxicity are not relevant to this situation. There is no unanimity of opinion among regulatory agencies as to the minimum survival required to produce a valid carcinogenicity study or as to the best approach for dealing with survival problems. Even with a single agency such as the FDA, different opinions exist on these issues. For example, the recently issued FDA Redbook II Draft Guidelines requires that rats, mice, or hamsters be treated for 24 months. Early termination due to decreased survival is not recommended. The EEC guidelines differ in that they suggest termination of the study when survival in the control group reaches 20%, while Japanese guidelines suggest termination at 25% survival in the control or low-dose groups (Speid et al., 1990). These provisions make good sense in that they do not request termination of the study when device-related mortality may be present only at the highest dose.

Parameters Evaluated In a pure carcinogenicity study the chief parameters measured are survival and occurrence of tumors (Table 14.1). Also measured are typically urinalysis parameters on samples collected prior to study start, at six month intervals during the study and just prior to the final sacrifice, as presented in Table 14.2. Clinical pathology and hematology measurements are made on blood samples collected at the same intervals with parameters measured in Table 14.2.

Statistical Analysis Irrespective of the specific protocols used, all carcinogenicity studies end with a statistical comparison of tumor proportions between treated and control groups. This analysis is

Lifetime Carcinogenicity Study (Implant) Organs and Sites to be Examined

Adrenals (2) Brain (3 levels: forebrain, mid and hindbrain including brainstem) Eyes (2) Gastrointestinal tract: Esophagus Stomach (glandular and nonglandular) Duodenum Jejunum Ileum Cecum Colon Rectum Gonads: ovaries with oviducts (2) testes with epididymids (2)

● ● ● ● ● ●

● ● ● ● ● ● ● ● ●

Harderian glands (2) Heart Kidneys (2) Larynx Liver (2 lobes) Lung (2 coronal sections including all lobes and mainstem bronchi) Lymph node (mesenteric) Mammary region (males and females) Pancreas Pituitary Prostate Salivary gland (submaxillary) Sciatic nerve Seminal vesicles Skeletal muscle (thigh)

● ● ● ● ● ● ● ● ● ●

●

Spleen Spinal cord (cervical) Skin (dorsal) Sternum/bone marrow Thymic region Thyroid/parathyroid Trachea Urinary bladder Uterus Implant site (4 sections of subcutaneous site and contiguous performal region) Any other grossly abnormal tissues or organs

200

Safety Evaluation of Medical Devices

Table 14.2 ● ● ● ● ● ● ● ● ● ● ●

Urinalysis Parameters Measured

Appearance (color) pH Ketones Urobilinogen Specific gravity (refractive index) Albumin Glucose Occult blood Urinary sediment Volume Bilirubin

necessary because the control incidence of most tumor types is rarely zero. In the unlikely case that a type of tumor is found in treated animals but not in concurrent or appropriate historical controls, it is reasonable to conclude that the tumor is treatment-related without statistical analysis (Haschek and Rousseaup, 1991). Most companies analyze tumor data using mortalityadjusted methods (PMA, 1988). Peto/International Agency for Research on Cancer (IARC) methodology is most commonly used, perhaps because this method is currently favored by the FDA (Peto et al., 1980). The use of life-table methods is most appropriate for “lethal” tumors, that is, those that cause the death of the animals. Various statistical methods are available for analyzing the incidence of lethal and nonlethal tumors (e.g., Gart et al., 1979, 1986; Dinse and Lagakos, 1983; McKnight, 1988; Portier and Bailer, 1989; Gad, 2004). These methods are especially useful when there are drug-related differences in mortality rates. When there is no drug effect of survival, unadjusted methods will generally give the same results. As a general approach, most pharmaceutical statisticians begin by testing for the presence of a dose-related trend in tumor proportions. If the trend test is significant, that is, the p value is less than or equal to 0.05, pairwise comparisons are performed between the treated and control groups. Trend and pairwise analyses may be adjusted for mortality as stated earlier, or performed without mortality adjustment using such simple methods as chi-square or Fisher’s exact tests. Although in most cases the use of trend tests is appropriate since most biological responses are dose-related, there are exceptions to this rule. Certain drugs, especially those with hormonal activity, may not produce classical dose responses and may even induce inverse dose–response phenomena. In these cases, a pairwise comparison may be appropriate in the absence of a significant positive trend. Most companies use one-tailed comparisons, and a substantial number use two-tailed methods. Since regulatory agencies are primarily interested in identifying carcinogenic drugs, as opposed to those that inhibit carcinogenesis, the use of one-tailed tests is generally considered more appropriate. Some companies prefer two-tailed comparisons because, in the absence of a true carcinogenic effect, there is an equal probability of seeing significant decreases as well as significant increases by chance alone. One of the most important statistical issues in the analysis of carcinogenicity data is the frequency of “false positives,” or type I errors. Because of the multiplicity of tumor sites examined and the number of tests employed, there is concern that noncarcinogenic devices may be erroneously declared carcinogens. If any p < 0.05 increase in tumor incidence is automatically regarded as a biologically meaningful result,

then the false positive rate may be as high as 47–50% (Haseman et al., 1986). Several statistical procedures designed to correct for the multiplicity of significance tests have been published (Haseman, 1990). One approach to the problem of multiple tumor site/type testing is a procedure attributed to Tukey by Mantel (1980). This method is used to adjust a calculated p value based on the number of tumor types/sites for which there are a minimum number of tumors in the particular study. The reasoning here is that, for most tumor sites, the number of tumors found is so small that it is impossible to obtain a significant result for that tumor site no matter how the tumors might have been distributed among the dose groups. Only those sites for which a minimum number of tumors is present can contribute to the false positive rate for a particular study. A method proposed by Schweder and Spjotvoll (1982) is based on a plot of the cumulative distribution of observed p values. Farrar and Crump (1988) have published a statistical procedure designed not only to control the probability of false positive findings, but also to combine the probabilities of a carcinogenic effect across tumor sites, sexes, and species. Another approach to controlling the false positive rate in carcinogenicity studies was proposed by Haseman (1983). Under this "rule," a compound would be declared a carcinogen if it produced an increase significant at the 1% level in a common tumor or an increase significant at the 5% level in a rare tumor. A rare neoplasm was defined as a neoplasm that occurred with a frequency of less than 1% in control animals. The overall false positive rate associated with the decision rule was found to be no more than 7–8%, based on control tumor incidences from NTP studies in rats and mice. This false positive rate compares favorably with the expected rate of 5%, which is the probability at which one would erroneously conclude that a compound was a carcinogen. This method is notable for its simplicity, and deserves serious consideration by pharmaceutical statisticians and toxicologists. Without resorting to sophisticated mathematics, this method recognizes the fact that tumors differ in their spontaneous frequencies and, therefore, in their contribution to the overall false positive rates in carcinogenicity studies. False positive results are much less likely to occur at tissue sites with low spontaneous tumor incidences than at those with high frequencies. As a final point that has special relevance to pharmaceutical carcinogenicity studies, one may question whether the corrections for multiple comparisons and their effect on the overall false positive rate are appropriate for all tumor types. For example, if a compound is known to bind to receptors and produce pharmacological effects in a certain organ, is it justified to arbitrarily correct the calculated p value for the incidence of tumors in that organ, using the methods described above? It is difficult to justify such a correction considering that the basis for correcting the calculated p value is that the true probability of observing an increased incidence of tumors at any site by chance alone may be much higher than the nominal alpha level (usually 0.05). It is reasonable to expect that, when a drug has known pharmacological effects on a given organ, the probability of observing an increased tumor incidence in that organ by chance alone is unlikely to be higher than the nominal 5% alpha level. Although most pharmaceutical statisticians and toxicologists agree on the need to control the probability of false positive results, there is no consensus as to which method is

Chapter 14:

most appropriate or most acceptable to regulatory agencies. The FDA and other such agencies will accept a variety of statistical procedures but will often reanalyze the data and draw their own conclusions based on their analyses.

INTERPRETATION OF RESULTS Criteria for a Positive Result There are three generally accepted criteria for a positive result in a carcinogenicity study. The first two are derived directly from the results of the statistical analysis: (1) a statistically significant increase in the incidence of a common tumor and (2) a statistically significant reduction in the time to tumor development. The third criterion is the occurrence of very rare tumors, that is, those not normally seen in control animals, even if the incidence is not statistically significant.

Use of Historical Controls When the study is over, the data analyzed, and the p values corrected, as appropriate, one may find that one or more tumor types increased in drug-treated groups relative to concurrent control group(s), but comparable to or lower than the historical incidence. Occasionally, a small number of tumors may be found in a treated group and the incidence may be significant because of the absence of this tumor in the concurrent controls. Review of appropriate historical control data may reveal that the low tumor incidence in the treated group is within the “expected” range for this tumor. The role of historical control data in interpreting carcinogenicity findings depends on the “quality” of the historical data. Ideally, the data should be derived from animals of the same age, sex, strain and supplier, housed in the same facility, and the pathology examinations should have been performed by the same pathologist or using the same pathological criteria for diagnosis. Since genetic drift occurs even in animals of a given strain and supplier, recent data are more useful than older data. The value of historical control data is directly proportional to the extent to which these conditions are fulfilled. Although methods are available for including historical control data in the formal statistical analysis (Tarone, 1982; Dempster et al., 1983), this is usually not done and for good reason. The heterogeneity of historical data requires that they be used qualitatively and selectively to aid in the final interpretation of the data, after completion of the formal statistical analysis.

REFERENCES Alexander, P. and Horning, E.S. (1959). Observations on the Oppenheimer method of inducing tumors by subcutaneous implantation of plastic films. In: G.E.W. Wolstenholme and M. O’Conner, (Eds.), Ciba Foundation Symposium on Carcinogenesis, Mechanisms of Action. Little Brown and Co., Boston, Massachusetts, pp. 12–25. Arnold, D.L., Grice, H.C. and Krawski, D.R. (1990). Handbook of In Vivo Toxicity Testing. Academic Press, San Diego. ASTM (1981). Performance of lifetime bioassay for the tumorigenic potential of implant materials, Designation F1439-92. ASTM Standards Book, Volume 13, pp. 761–765. ASTM, Philadelphia, PA. Bates, R.R. and Klein, M. (1966). Importance of a smooth surface in carcinogenesis by plastic film. J. Natl. Cancer Inst., 37: 145–151. Bischoff, F. and Bryson, G. (1964). Carcinogenesis through solid state surfaces. Prog. Exp. Tmor. Res., 5: 85–113. Brand, G.G., Johnson, K.H. and Buoen, L.C. (1976). Foreign Body, Tumorigenesis CRC Crit. Rev. In: Toxicology, October, p. 353.

Carcinogenicity

201

Brand, K.G., Buoen, L.C., Johnson, K.H. et al. (1975). Foreign body in foreign body tumorigenesis: a review. Cancer Res., 35: 279–286. Brand, L. and Brand, K.G. (1980). Testing of implant materials for foreign body carcinogenesis. In: G.D. Winter, D.G. Gibbons and H. Plenk Jr. (Eds.), Biomaterials, 1980, p. 819. Advances in Biomaterials, Vol. 13. J. Wiley, New York: (1982). Cameron, T.P., Hickman, R.L., Kornreich, M.R. and Tarone, R.E. (1985). History, survival, and growth patterns of B6C3F1 mice and F344 rats in the National Cancer Institute Carcinogenesis Testing Program. Fundam. Appl. Toxicol., 5: 526–538. Chu, K., Cueto, C. and Ward, J. (1981). Factors in the evaluation of 20 NCI carcinogenicity bioassays. J. Toxicol. Environ. Health, 8: 251–280. Dempster, A.P., Selivyn, M.R. and Weeks, B.J. (1983). Combining historical and randomized controls for assessing trends in proportions. J. Amer. Stat. Assoc., 78: 221–227. DePass, L.R., Weil, C.S. and Ballentyne, B. (1986). Influence of housing conditions for mice on results of a dermal oncogenicity bioassay. Fundam. Appl. Toxicol., 7: 601–608. Dinse, G.E. and Lagokos, S.W. (1983). Regression analysis of tumor prevalence data. Appl. Stat., 32: 236–248. Dukes, C.E. and Mitchley, B.C.V. (1962). Polyvinyl sponge implants: experimental and clinical observations. Br. J. Plast. Surg., 15: 225–235. Ecanow, B., Gold, B.H. and Sadove, M. (1977). The role of inert foreign bodies in the pathogenesis of cancer. Br. J. Cancer., 36: 397. Farrar, D.B. and Crump, K.S. (1988). Exact statistical tests for any carcinogenic effect in animal bioassays. Fundam. Appl. Toxicol., 11: 652–663. FDA (2001). Guidance for Industry: Statistical Aspects of the Design, Analysis and Interpretation of Chronic Rodent Carcinogenicity Studies of Pharmaceuticals, VSDHEW, Washington, DC. Fox, J.G. (1977). Clinical assessment of laboratory rodents on long term bioassay studies. J. Environ. Pathol. Toxicol., 1: 199–226. Gad, S.C. (2002). Animal Models in Toxicology, 2nd Edn. CRC Press, Boca Raton, FL. Gad, S.C. (2004). Statistics and Experimental Design for Toxicologists, 4th Edn. Taylor & Francis, Boca Raton, FL. Gart, J.J., Chu, K.C. and Tarone, R.E. (1979). Statistical issues in interpretation of chronic bioassay tests for carcinogenicity. J. Natl. Cancer Inst., 62: 957–974. Gart, J.J., Krewski, D., Lee, P.A., Tarone, R.E. and Wahrendorf, J. (1986). Statistical Methods in Cancer Research, Vol. III. The Design and Analysis of Long-term Animal Experiments. IARC Scientific Publication No. 79. International Agency for Research on Cancer, Lyon. Goldhaber, P. (1961). The influence of pore size on carcinogenicity of subcutaneously implanted Millipore filters. Proc. AM. Assoc. Cancer Res., 3: 228. Goldhaber, P. (1962). Further observations concerning the carcinogenicity of Millipore filters. Proc. Am. Assoc. Cancer Res., 3: 323. Hajian, G. (1983). Statistical issues in the design and analysis of carcinogenicity bioassays. Toxicol. Pathol., 11: 83–89. Haschek, W.M. and Rousseaup, C.G. (1991). Handbook of Toxicology Pathology. Academic Press, San Diego. Haseman, J.K. (1983). A reexamination of false-positive rates for carcinogenicity studies. Fundam. Appl. Toxicol., 3: 334–339. Haseman, J.K. (1990). Use of statistical decision rules for evaluating laboratory animal carcinogenicity studies. Fundam. Appl. Toxicol., 14: 637–648. Haseman, J.K., Winbush, J.S. and O’Donnell, M.W. (1986). Use of dual control groups to estimate false positive rates in laboratory animal carcinogenicity studies. Fundam. Appl. Toxicol., 7: 573–584. Henry, T.J. (Ed.) (1985). HIMA Report 85-1, Guidelines for the Preclinical Safety Evaluation of Materials Used in Medical Devices. Washington, D.C., Health Industry Manufacturers Association. International Agency for Research on Cancer (1987). IARC Monographs on the Evaluation of Carcinogenic Risks to Humans—Preamble. IARC Internal. Technical Report 87/001, IARC, Lyon. ISO (1995). ISO10993—3: Carcinogenicity.

202

Safety Evaluation of Medical Devices

Littlefield, N.A., Farmer, J.H., Taylor, D.W. and Sheldon, W.G. (1979). Effects of dose and time in a long term, low-dose carcinogenicity study. In: J.A. Staffer and M.A. Shellman (Eds.), Innovations in Cancer Risk Assessment (EDOI Study). Pathlon Publishers, FL, pp. 17–34. Mantel, N. (1980). Assessing laboratory evidence for neoplastic activity. Biometrics, 36: 381–399. McKnight, B. (1988). A guide to the statistical analysis of long-term carcinogenicity assays. Fundam. Appl. Toxicol., 10: 355–364. Memol, V. (1986). Malignant neoplasms associated with orthopedic implant materials in rats. I Orthoped Res., 4: 346–355. Oppenheimer, B.S., Oppenheimer, E.T. and Stout, A.P. (1948). Sarcomas induced in rats by implanting cellophane. Proc. Soc. Exp. Biol. Med., 67: 33–34. Oppenheimer, B.S., Oppenheimer, E.T. and Stout, A.P. (1952). Sarcomas induced in rodents by imbedding various plastic films. Proc. Soc. Expt. Biol. Med., 79: 366–369. Oppenheimer, B.S., Oppenheimer, E.T., Stout, A.P. et al. (1953). Malignant tumors resulting from embedding plastics in rodents. Science, 118: 305–306 Oppenheimer, B.S., Oppenheimer, E.T., Danishefsky, I. et al. (1955). Further studies of polymers as carcinogenic agents in animals. Cancer Res., 15: 333–340. Oppenheimer, B.S., Oppenheimer, E.T., Stout, A.P. et al. (1958). The latent period in carcinogenesis by plastic in rats and its relation to the presarcomatous stage. Cancer, 11: 204–213. Oppenheimer, E.T., Willhite, M., Danishefsky et al. (1961). Observations on the effects of powdered polymer in the carcinogenic process. Cancer Res., 21: 132–136. Oppenheimer, E.T., Willhite, M., Stout, A.P. et al. (1964). A comparative study of the effects of imbedding cellophane and polystyrene films in rats. Cancer Res., 24: 379–387.

Peto, R., Pike, M.C., Day, N.E., Gray, R.G., Lee, P.N., Parish, S., Peto, J., Richards, S. and Wahrendorf, J. (1980). Guidelines for Simple, Sensitive Significance Tests for Carcinogenic Effects in Long-Term Animal Experiments. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. International Agency for Research on Cancer, Lyon. PMA (1988). Results of a Questionnaire Involving the Design and Experience with Carcinogenicity Studies. Pharmaceutical Manufacturers Association (now PhRMA), Washington, D.C. Portier, C.J. and Bailer, A.J. (1989). Testing for increased carcinogenicity using a survival-adjusted quantal response test. Fundam Appl. Toxicol., 12: 731–737. Rao, G.N. and Huff, J. (1990). Refinement of long-term toxicity and carcinogenesis studies. Fundam. Appl. Toxicol., 15: 33–43. Sher, S.P., Jensen, R.D. and Bokelman, D.L. (1982). Spontaneous tumors in control F344 and Charles River CD rats and Charles River CD-1 and B6C3F1 mice. Toxicol. Lett., 11: 103–110. Speid, L.H., Lumlez, C.F. and Walker, S.R. (1990). Harmonization of guidelines for toxicity testing of pharmaceuticals by 1992. Res. Toxicol. Pharmacol., 12: 179–211. Schweder, T. and Spjotvoll, E. (1982). Plots of p-values to evaluate many test simultaneously. Biometrika, 60: 493–502. Tarone, R.E. (1982). The use of historical control information in testing for a trend in proportions. Biometrics, 38: 215–220. Turner, F.C. (1941) Sarcomas at sites of subcutaneously implanted Bakelite disks. J. Nat. Cancer Inst., 2: 81–83. Zbinden, G. (1993). The Concept of Multispecies Testing in Industrial Toxicology, Regul. Toxicol. Pharmacol., 17: 84–94.

15 Degradation Products and Impurities

All materials, including metals, ceramics, and polymers, are subject to biodegradation. The major degradative processes specific to each type of material, and methods for assessment of the extent of degradation, are discussed below.

INTRODUCTION The human body is a complex milieu that subjects medical devices to extreme physical and chemical challenges. Implanted materials may be subjected to very high stresses while under attack by a corrosive environment and the body’s immune system. Given these conditions, it is not surprising that these materials often suffer degradation that may, in the best case, produce harmless byproducts, or in the worst case, lead to device failure, disease, or death. The intracellular and extracellular body fluids contain multiple organic and inorganic compounds. Extracellular fluids including blood, urine, interstitial fluid, digestive fluids, and others, contain salts, metals, amino acids, and proteins. They may also include cells or parts of cells (Tables 15.1 and 15.2) The pH of body fluids is about 7.4, although it may change temporarily in response to inflammatory processes associated with surgical implantation procedures. Body temperature is constant at approximately 37°C. Table 15.1

METALS Corrosion Metallic corrosion inevitably occurs when metals are placed in contact with physiological tissues and fluids, and may result in the release of undesirable metal ions or products that can lead to profound local tissue damage and degradation of devices. Systemic responses may produce long lasting effects, e.g., hypersensitivity, or in worst case scenarios, result in loss of life. Commonly observed forms of corrosion include pitting/crevice, fretting, stress, fatigue, and galvanic corrosion. The process of corrosion, shown in Figure 15.1, is a function of both the chemical composition of the metal or alloy and the complex chemistry presented by the physiological environment. Two reactions that always occur when metals are placed in an aqueous electrolyte are:

Ionic Concentrations in Blood Plasma and Extracellular Fluid

Anion/cation

Blood plasma (mM)

−

Cl HCO3− HPO42− SO42− H2PO4− Na+ Mg2+ Ca2+ K+

96–111 16–31 1–1.5 0.35–1 2 131–155 0.7–1.9 1.9–3 35–1.6

Extracellular fluid (mM) 112–120 25.3–29.7 193–102 0.4 — 141–15 1.3 1.4–1.55 3.5–4

1. An anodic reaction that yields metallic ions and electrons. 2. A cathodic reaction in which the electrons are consumed. The rates of these two reactions are always equal. Perturbations in the local environment that affect the rate of either reaction and thereby shift the balance may determine the course of the ensuing corrosive process. Corrosion can be prevented by suppression of either of these reactions. The potential difference between a metal and a standard reference electron in solution, the electrochemical potential, is a general indicator of the likelihood of a metal to corrode. Noble metals such as gold, platinum, and silver, are high on the electrochemical scale, and are unlikely to corrode. On the other end of the scale, chromium, aluminum, and titanium have much higher reactivity, and would be expected to corrode readily in body fluids, as can be seen in Table 15.3. In an aqueous solution that does not contain its own ions, the Nernst equation predicts the electrode potential at which the rate of metal dissolution is equal to the rate of the cathodic reaction. Presumably, the reaction will occur until this potential is reached and the system is in equilibrium. Unfortunately, this simple model does not account for inhomogeneities in the implant environment, or the other factors that result in displacement of the equilibrium, or “polarization.” Metal complexation with proteins, and subsequent migration of those proteins, has been shown to reduce the

Source: Bundy (1994). Table 15.2 Major Proteins and Other Organic Constituents of Blood Plasma Constituent Albumin α-Globulins β-Globulins γ-Globulins Alpho-lipoproteins Fibrinogen Total cholesterol Fatty acids Glucose Lactate Urea

Concentration (g/L unless otherwise stated) 30–55 5–10 6–12 6.6–15 3.5–4.5 1.7–4.3 1.2–2.5 1.9–4.5 0.65–1.1 0.5–2.2 mM 3–7 mM

Source: Bundy (1994).

203

204

Safety Evaluation of Medical Devices

Electrolyte Solution M+ +OH-

Corrosion Product

OH-

M+

M+

M+

M+

OH-

OH-

OH-

OH-

M+ eeeee-

Anode Region

Cathode Region

Figure 15.1. Schematic of corrosion process in metals. Metallic surfaces are divided into small oxygen deficient and oxygen rich regions. These become the anode and cathode regions respectively. Electrons (e−) diffuse from the anode into the cathode, leaving positively charged metal ions (M+). The electrons then form negatively charged hydroxide ions in the electrolyte solution (OH+). These hydroxide ions combine with the metal ions to yield a corrosion product, such as rust (Fe(OH)2) in the case of iron.

concentration of metal ions in the interface region, forcing further release of metal ions to maintain the equilibrium. Relative movement at the implant–tissue interface results in local disruption of the electrolyte composition that shifts the balance toward corrosion. The extent of corrosion that occurs is determined by the nature of the polarization, and the ability of the metal to respond to it. Metallic corrosion is a surface effect. Since most nonreactive metals, e.g., the noble metals, are not suitable for load bearing applications, surface modification of more reactive metals is a useful strategy for reducing corrosion while maintaining favorable bulk properties. Reactive metals, e.g., aluminum, titanium, and chromium, react with gaseous oxygen to form metal oxide layers by a process known as passivation. The metal oxide layers act as a barrier to diffusion of metal ions, and therefore, prevents further corrosive reactions at the surface. Damage to this layer in vivo may greatly accelerate the degradation process. Surface interactions with proteins or bacteria that alter local pH or reduce the availability of oxygen may compromise the stability of Table 15.3

Electrochemical Scale

Reactivity Noble

Active

Reaction Au + 3e Pt2− + 2e− Ag+ + e− O2 + 2H2O + 4e− Ti(OH)3+ + H+ + e− H+ + e− Fe3+ + 3e− Co2+ + 2e− Fe2+ + 2e− Cr2+ + 2e− Cr3+ + 3e− 2H2O + 2e− TiO2 + 4H+ + 4e− Ti2+ + 2e− Mg− + e− Na+ + e− 3+

−

Source: Jacobs et al. (1998).

Potential (V)

⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔ ⇔

Au Pt Ag OH− Ti3− + H2O 1/2 H2 Fe Co Fe Cr Cr 2OH− Ti + 2H2O Ti Mg Na

1.42 1.20 0.80 0.40 0.06 0.00 −0.04 −0.28 −0.41 −0.56 −0.74 −0.83 −0.86 −1.60 −2.37 −2.71

an oxide layer. It should be noted that even an intact oxide layer will allow some diffusion of metal ions. Metallic implant materials are most commonly evaluated in phosphate buffered saline (PBS), Hank’s solution, Ringer’s solution, or artificial saliva (for dental applications). Hank’s solution and Ringer’s solution contain physiologically relevant salts. Recent changes to ASTM F2129 (ASTM International, 2006) specify a PBS that does not include calcium or magnesium (both of which are found in Hank’s and Ringer’s solutions) since these tend to precipitate from the solutions and affect the pH. Numerous compositions for artificial saliva are given in Table 15.4. Since protein content is known to significantly affect the corrosiveness of body fluids, recent in vitro studies have included organic compounds in the test solutions. Pitting is initiated by a localized breakdown of the protective oxide layer that results in localized chemical corrosion, and propagates through high rate anodic dissolution within an occluded cavity at potentials where most of the surface remains passivated (Moayed and Newman, 2006). The pit acts as the anode, while the rest of the surface acts as the cathode. Since the rates of anodic and cathodic reactions must be equal, a high degree of damage may be incurred. Epit, the “pitting potential,” is the potential above which stable pit propagation can occur. Crevice corrosion occurs in narrow openings between metals or between metals and non-metals, where there is localized oxygen depletion that accelerates corrosion. Some authors consider pitting to be a special form of crevice corrosion, and others feel that crevice corrosion is a special form of pitting. This type of corrosion is commonly seen in porous coated implants, and at the interface of interlocking components in orthopedic implants. Fretting corrosion occurs in response to repetitive scratching or mechanical disruption of the passivated surface. Factors that contribute to fretting include the removal of the oxide film, mechanical stress resulting in surface deformation that can in turn lead to fatigue corrosion, and perturbation/mixing of the electrolyte at the interface. The combination of stress and a corrosive environment lead to the phenomenon known as stress corrosion. This type

Chapter 15:

Table 15.4 Ca2+ 5.80 0 0 0 1.43 1.5 1.0

Degradation Products and Impurities

205

Compositions of Artificial Saliva (Mmol/L) CO3 0 7.10 0 17.85 6.45 15.00 17.86

Mg2+ 0 0 0 0 0 0 0

K+

Na+

C1−

SCN−

P

NH4+

pH

58.60 0 25.74 21.48 20.5 0 0

14.64 27.14 28.16 33.49 11.55 37.5 51.14

32.19 29.80 29.84 28.07 23.22 3.00 37.08

0 2.5 2.24 3.4 2.3 0 0

16.79 4.70 4.7 2.98 5.1 15.00 1.0

0 4.10 4.1 0 4.35 0 0

6.7 6.8 — 6.7 — — —

Source: Singh and Dahotre (2007).

of corrosion is initiated at a microscopic defect or crack that does not immediately repassivate, and grows incrementally until the crack results in device failure. A second type of corrosion, known as fatigue corrosion, occurs through repetitive or cyclic stressing in the corrosive environment. Galvanic corrosion occurs when two metals are put in contact with one another in an electrolyte. It can also occur at the grain boundaries in an alloy. A difference in the electrochemical potentials of the metals causes electrons to flow from the more anodic metal to the more cathodic metal, thereby selectively degrading the anodic metal while the cathodic metal is protected.

Corrosion Testing Several standardized test methods are used to evaluate the corrosion resistance of metals, including the original test method, ASTM G 5, a potentiodynamic test of electrochemical potential, and ASTM G 61, a cyclic experiment used to determine susceptibility to pitting and corrosion. In the late 1970s, researchers at Stanford Research Institute developed a novel method for assessing the pitting resistance of stainless steel alloys using passive–pitting– repassivation (PPR) curves (Syrett, 1977). The method was eventually modified to create ASTM Standard F 746, which ranks the resistance of alloys to pitting under specific conditions designed to induce corrosion. The results of the test did not always predict in vivo corrosion of implants made from these alloys. The test involved placing a nonmetallic collar around a specimen immersed in a 0.9 NaCl solution for one hour while measuring the potential with respect to a standard electrode. After one hour, the potential of the specimen was shifted to +0.8 V and held for a specific period of time, while the resultant current changes were monitored. Finally, the potential was shifted back to the original level, and the current was recorded to assess repassivation. Since materials processing that occurs during implant formation and finishing can greatly affect surface properties, new test methods were designed to assess finished devices in their ready-to-implant states. ASTM F 2129 assesses small metallic devices or components using the G 61 technique. The devices are immersed in deaerated PBS and the resting potential is measured. After one hour, a potentiodynamic polarization scan is made, and then reversed when the current reaches a level two decades greater than that of the breakdown potential, and is stopped when the current becomes less than in the forward scan. This test is used for small implants, e.g., vascular stents, surgical staples, etc. While the FDA requires device manufacturers to demonstrate that their devices will not corrode within the expected lifetime, there is no acceptance criterion specified for any of the metal corrosion standards. It is common practice for device manufacturers to compare new product

data to data for products that are already on the market. The limitations of electrochemical testing must be considered by device manufacturers in prediction of product lifetimes.

Wear Despite the high level of strength and hardness of metal alloys commonly used in implantable devices, some wear does occur. Metallic wear is enhanced by a coupled mechanical/electrochemical degradative process known as tribocorrosion (Yan, 2006). Wear can be generated in response to intended motion of apposing bearing surfaces, or by unintended interactions between nonbearing surfaces or with third-body particles (bone, cement, or metal) (Jacobs et al., 1998). The prevalence of particulate disease resulting from wear of polymer bearing surfaces in total joint arthroplasty has renewed interest in the use of metal-on-metal (MM) implants (Brown, 2007). There have been numerous attempts to characterize metallic wear particles from periprosthetic tissues (Brown, 2006, 2007; Catelas et al., 2001; Doorn et al., 1998; Firkins et al., 2001) Wear particles collected from patients with MM total hip replacements (THRs) show a range of sizes and compositions, with mean sizes of less than 50 nm (Brown, 2007). Implantation time appeared to affect the size, shape, and composition of the particles. It is not known if these differences are a result of changes in the general quality of the bearings over the past 20 years, or if they are characteristic of the implantation duration. Inflammatory responses in the tissue surrounding MM THRs appear to be similar but less intense than those typically seen in response to polyethylene wear. Tissue responses to wear particles generated by early MM prostheses included necrosis and necrobiosis (Doorn et al., 1996). Systemic effects of MM wear are cause for concern. Low concentrations of metallic wear particles have been found in post-mortem specimens removed from the lymph nodes, livers, and spleens of patients with primary and revision joint replacements (Case et al., 1994; Jacobs, 2006; Urban et al., 2004). These particles are associated with degradation of both bearing and non-bearing surfaces. There is growing evidence supporting the cytotoxic effects of metal ions generated by corrosion of high surface area metallic wear particles. Serum levels of chromium, nickel, and cobalt ions in patients with failed hip implants were significantly higher than those in patients with no implants (Savarino et al., 1999). Both failed and intact THRs have also been associated with reductions in serum levels of white blood cells (Savarino et al., 1999). Numerous investigators have demonstrated significant decreases in cell viability upon direct contact with metal particles in vitro (Brown, 2006).

206

Safety Evaluation of Medical Devices

Metallic wear has also been implicated in hypersensitivity reactions including dermatitis, urticaria, and vasculitis. The prevalence of dermal sensitivity in patients with a total joint replacement is substantially higher than that in the general population (10 –15%). Metal hypersensitivity may be either an immediate humoral response, or a delayed cellmediated response; implant hypersensitivity responses are generally the latter type (type IV or delayed type hypersensitivity (DTH)). Type IV responses involve the activation and clonal expansion of T-type lymphocytes by antigen-presenting cells (APCs). Metal ions form complexes with native proteins that can act as antigens that stimulate an immune response. While the exact mechanism of the transport and delivery of metal ions to APCs are not yet known, the consequences are that, upon presentation of APCs, metal-ion sensitized T-lymphocytes release cytokines that recruit and activate macrophages. Historically, testing for DTH has been conducted by skin patch testing. Dermal exposures are shorter in duration (48–96 hr) and involve different effector cells (Langerhans) than those experienced in a long-term implantation. It is now recognized that skin patch testing is not always an indication of the potential for DTH. Alternative test methods include in vitro proliferation testing (also known as lymphocyte transformation testing, or LTT), and in vitro leukocyte migration inhibition testing (Brown, 2006; Hallab et al., 2001). LTT measures the proliferative response of lymphocytes following activation. While this method has been wellestablished for evaluation of metal sensitivity in other clinical applications, there have been few studies using this method for assessment of metal implant related hypersensitivity. In vitro leukocyte migration inhibition testing involves quantification of the rate of migration of leukocytes through or along media (capillary tube, membrane filter, collagen gel, agarose droplets, or layers) toward a known chemo-attractant. In the presence of sensitizing agent, leukocytes typically show a reduced ability to migrate toward chemo-attractants. The in vitro test methods have been underutilized in the assessment of implant-induced hypersensitivity. Improvement of methods, including lymphocyte transformation testing, migration inhibition, and cytokine enzyme-linked immunosorbent assay (ELISA) methods is expected to enhance the capability to correlate humoral responses with clinical conditions (Hallab et al., 2001). It is well established that chromium ions have a genotoxic effect in humans. Intracellular reduction of Cr(VI) to Cr(III) yields reactive species that can damage DNA and organelles. It has been difficult to establish a clear relationship between metal THRs and the incidence of cancer due to other possible contributory factors, including drug therapy and radiation (Case et al., 1996; Visuri et al., 1996). Assessment of the long-term effects of the systemic distribution of metallic wear particles will require further research. Generation of these potentially harmful agents can be reduced by improvements in implant design, use of proper implantation techniques, and removal of failing or corroding devices.

CERAMICS AND GLASSES Bioceramics have been used in multiple applications, including dental fillings, orthopedic implants, and tissue engineering substrates. Although some ceramics dissolve readily in an

aqueous environment, ceramics and glasses are generally known for their excellent stability and corrosion resistance under normal use conditions. Alumina (Al2O3) and zirconia (ZrO2) have been used in THRs as the femoral head component in combination with ultrahigh molecular weight polyethylene (UHMWPE) cups, and alumina has been used as both head and cup in ceramicon-ceramic (CC) implants. Wear rates for alumina on UHMWPE have been reported to be as much as 20 times lower than those of metal on UHMWPE (Katti, 2004). Bioresorbable calcium phosphate ceramics, e.g., tricalcium phosphate (TCP) (Ca3(PO4)2) and hydroxyapatite (HAP)(Ca10(PO4)6(OH)2), have been used extensively as metallic implant coatings and bone defect fillers due to their biocompatibility and osteoconductivity. Their uses are limited due to difficulty in fixation, low strength, and limited fatigue resistance. Bioresorbable ceramic implant materials are designed to break down into naturally occurring constituents, e.g., calcium and sodium compounds. Bioactive ceramics, or bioglasses, react with ions in polyelectrolyte body fluids to form surface silanol groups, which then undergo condensation to form a highly hydrated gel. Ca2+ and PO43− react at the gel surface to form a crystalline hydroxycarbonate apatite layer (CHA) that is compatible with the mineral phase of bone (Kenny and Buggy, 2003).

Degradation There have been numerous studies that demonstrate that degradation of calcium phosphate biomaterials in vivo occurs by: ●

●

Disintegration into small particles that are phagocytized or transported to neighboring tissues, including the lymph nodes Dissolution in interstitial body fluids (Klein et al., 1983; Lu et al., 2002).

The rate, completeness, and mode of degradation have been found to be dependent upon chemical composition, crystallization mechanism, microstructure and porosity (Bohner, 2000; Lu et al., 2002). Sintered bioceramics (HA and β-TCP) processed at a high temperature are primarily degraded by extracellular liquid dissolution. Ca–P cement is formed by a physicochemical reaction that yields low crystallization, and is prone to disintegration and degradation through a dissolution process associated with a cellular process.

Wear As a general rule, ceramic components used in orthopedic implants demonstrate superior wear resistance as compared with metal and polymer components. A recent 5-year postoperative follow-up study on patients younger than 40 who received alumina-on-alumina THRs showed that wear was undetectable in 24 hips, where differentiation of the femoral head from the cup was possible on radiographs and no osteolysis was observed (Yoo et al., 2006). Y-TZP (yttria-stabilized tetragonal zirconia polycrystal) zirconia ceramic femoral head components were first introduced in 1988 as a tougher, less brittle alternative to alumina. To-date, over 600,000 zirconia femoral heads have been implanted (Chevalier, 2007). Zirconia’s outstanding crack resistance is a consequence of phase transformation toughening, a stress-induced phase transformation of metastable grains to a monoclinic phase at the crack tip, resulting in volume expansion and a compressive stress that acts to resist

Chapter 15:

crack propagation (Chevalier, 2006, 2007). Unfortunately, this metastability makes zirconia susceptible to aging in the presence of water (Figure 15.2). It was initially believed that this aging phenomenon would not occur at physiologically relevant temperatures, until approximately 400 Prozyr® femoral heads fractured within a very short time period in 2001 (US FDA). These failures have been attributed to an unusual aging process in two batches of femoral heads. The zirconia aging process is initiated by a slow surface transformation to the stable monoclinic phase in the presence of water or moisture vapor. For hip implants, this “surface” transformation may occur at the polished wearing surface or

Degradation Products and Impurities

207

at the hipstem–ball interface. The increase in volume that accompanies the transition leads to microcracking that allows water to penetrate further into the implant, thereby initiating a chain reaction that results in the generation of wear particle debris. There are conflicting reports regarding wear and osteolysis around zirconia implants. The differing clinical experiences suggest differences in processing that may affect degradation resistance. Surface treatments (e.g. cementation and nitridation) may help prevent aging while allowing retention of the good mechanical properties of zirconium. Additionally, new zirconia reinforced alumina composites take advantage of the

H2O

H2O

H2O

H2O

H2O

H2O

(a)

(b)

(c)

Surface Uplift

Figure 15.2. Schematic of the ageing process occurring in a zirconia cross-section, showing the transformation of neighboring grains.Transformed grains are gray. The black path represents the penetration of water due to microcracking around the transformed grains. (a) Nucleation on a particular grain at the surface, leading to microcracking and stresses on neighboring grain. (b) Growth of the transformed zone, leading to extensive microcracking and surface roughening. (c) Continued transformation and penetration with significant surface uplift. (Source: Adapted from J. Chevalier, Biomaterials, 2006, 27, 4, 535. With permission.)

208

Safety Evaluation of Medical Devices

toughening and crack resistance offered by the zirconia transformation while reducing the deleterious effects of aging.

POLYMERS Degradation Polymers degrade by a variety of physical and chemical mechanisms. Multiple processes in the total life cycle of a polymer-based medical device that may predispose the polymer to specific degradation mechanisms include: ● ● ● ● ● ● ● ●

Chemical synthesis Extrusion or molding Machining Packaging Sterilization Shipment Storage Deployment/implantation.

Degradation may begin at any stage prior to the intended device use, and can be accelerated by the environment at any stage, including implantation or exposure to body fluids. Polymers that contact tissue or body fluids, and particularly implanted devices, are subjected to chemical degradation via hydrolysis, oxidation, and enzymatic processes. The superposition of mechanical stress upon this chemical attack may exacerbate the degradation process. It should be noted that bioerosion of a solid polymer may occur in the absence of hydrolysis as a result of solubilization of the polymer.

Hydrolysis Hydrolysis is a reaction of polymer functional groups with water molecules that results in chain cleavage. Functional groups that are highly susceptible to hydrolysis include esters, amides, urethanes, carbonates, and anhydrides. Other polymer functional groups that may hydrolyze under certain conditions include ether, acetal, nitrile, phosphonate, sulfanate, sulfonamide, and methylenes. Groups that are normally stable to hydrolysis include hydrocarbons, halocarbons, dimethylsiloxane, and sulfone (Ratner, 2004). Since hydrolysis is dependent upon water reaching these functional groups, factors that affect water absorption, including crystallinity, cross-linking, porosity, and molecular weight all play a role in determining the rate and extent of hydrolytic degradation. Additionally, hydrolysis may be catalyzed by ions in the extracellular fluids, enzymes, or by reaction products (autocatalysis). Unintended hydrolytic degradation of polymers has significant impacts on both the expected lifetime and the biocompatibility of medical devices. Chain scission results in changes in the mechanical properties of the polymer, and the release of potentially cytotoxic byproducts. The in vivo hydrolytic degradation of commonly used synthetic polymers, including polyesters (PET), polyurethanes, cyanoacrylates, and polyamides has been extensively characterized. Highly crystalline aromatic polyesters are hydrophobic, and are generally considered to be hydrolytically stable. Polyamides (nylon 6 and 6,6) are hydrophilic, and as such, are subject to rapid hydrolysis that results in significant degradation and loss of tensile strength within weeks of implantation. Cyanoacrylate polymers have been shown to suffer extensive degradation in vivo within 4–6 months of implantation.

Polyester urethanes were among the earliest implanted polymers. Early studies showed that these materials degraded and fragmented within months of implantation. Polyester urethanes synthesized from toluene diisocyanate release toluenediamine (TDA), a known carcinogen, upon hydrolyis. Despite these concerns, use of polyester urethane foams as breast implant coatings persisted until 1991, when the manufacturer withdrew the implants from the market. These polyester urethanes were originally manufactured for use as filtration products, and were never tested for human implantation (Ericsson., 1998). TDA has been found in the breast milk and urine of women with polyurethane coated silicone implants (Chan et al., 1991; Heggers et al., 1983). In addition, there have been reports of two cases of complete “dissolution” of these implants (Black, 1991).

Resorbable Polymers Controlled degradation and erosion of hydrolytically unstable polymers is desirable in many medical devices, including sutures, tissue engineering scaffolds, and controlled drug delivery systems. Numerous synthetic polymers have been investigated as degradable/resorbable materials, including aliphatic polyesters, polyanhydrides, polycyanoacrylates, polyamino and pseudo-polyamino acids, polyorthoester, polyphosphazenes, polypropylene, and polypropylene fumarate. The requirement of nontoxic hydrolysis products imposes limitations on the selection of monomers. Addition of hydrolysable cross-links can be used to facilitate hydrolytic degradation of nonhydrolyzable polymer. The rate of erosion of these polymers is dependent on not only the reactivity of the hydrolyzable bond, but also the hydrophilicity and morphology of the bulk polymer (Gopferich, 1996). Crystalline regions are less easily penetrated, and therefore, amorphous regions degrade more rapidly. The size, shape, and number of crystallites affect chain mobility and degradation rate. Hydrophilicity is a function of the monomer chemistry, and therefore degradation of hydrolyzable polymers can be retarded by selection of less hydrophilic monomer. For example, polylactic acid is more hydrophobic than polyglycolic acid, and degrades more slowly, despite similar reactivities of the ester linkage with water. Copolymerization of hydrophilic and hydrophobic monomers is another strategy that can be used to alter the degradation rates with respect to the homopolymer. Other factors that may affect the degradation mechanism and rate include the incorporation of drugs or plasticizers, sterilization, and storage conditions (Shikanov et al., 2005). Biodegradable polymers erode by either surface or bulk erosion (Gopferich, 1996). In surface erosion, the rate of bond hydrolysis is much higher than the rate of diffusion of water into the bulk, and erosion occurs from the outside in, so that the overall volume of the polymer is decreased over time. Bulk erosion occurs when the rate of diffusion of water is much higher than the rate of bond hydrolysis. In bulk erosion, the polymer degrades rather uniformly throughout, and becomes porous as small degradation products diffuse out, eventually crumbling or cracking. The thickness of the material plays a role as well, such that a very thin sample of surface eroding polymer appears to degrade by bulk erosion, and a very thick sample of bulk eroding polymer appears to degrade by surface erosion (von Burkersroda et al., 2002). Evaluation of biodegradation of resorbable polymers is usually accomplished through a combination of in vivo and in vitro experiments, by retrieving samples at predetermined intervals and characterizing physical and

Chapter 15:

mechanical properties. In vitro studies usually involve submersion in a neutral PBS and incubation at 37°C, as prescribed by ISO 15814:1999 (Implants for surgery—Copolymers and blends based on polylactide—In vitro degradation testing) (ISO, 1999). A recent study comparing degradation of identical PLLA in vivo and in vitro found that in vitro degradation occurs at a much higher rate (Landes et al., 2006). Since resorbable polymers have been shown to persist in vivo for periods in excess of five years, a reliable accelerated in vitro degradation test method is highly desirable. Attempts to accelerate in vitro degradation have included application of stress, increasing temperature, and varying the pH. A recent investigation comparing in vitro degradation of poly-L-lactide at elevated temperatures (50–70°C) to in vivo degradation indicated that the degradation mechanisms were very similar to the in vitro and in vivo degradation mechanisms reported at 37°C (Weir et al., 2004). These results must be interpreted with caution due to concerns regarding the validity of testing above the Tg (66°C for initial poly-L-lactide, although the Tg should decrease due to plasticization with water). In vivo animal studies reflect species specific degradation rates and mechanisms. Self-reinforced PLGA 80:20 copolymer screws in rabbit cranial bone required more than 18 months to resorb, while in minipigs, complete degradation of 82:18 PLGA was observed after 12 to 18 months (Tiainen et al., 2004; Wiltfang et al., 2000). In chinchilla rabbit femurs, complete degradation of PLGA implants occurred after 14 months (Viljanen et al., 2001). A 2006 study characterized the tissue response of rabbit femur cancellous bone to polyglycolide (PGA), polydioxanone (PDS), polylevolactide (PLLA), and stainless steel pins under identical conditions. PGA and PDS pins degraded completely within 12 months, with no visible degradation of the PLLA within that time period (Pihlajamaki et al., 2006). A recent in-patient evaluation estimates that complete degradation of PLGA in human patients will reliably occur at about 12 months. Comparison of resorption rates in animals and humans provides evidence that hydrolytic degradation is predominant in humans, while cellular enzymatic processes play a more prominent role in animals (Heidemann et al., 2003; Tiainen et al., 2004).

Oxidation Polymers that degrade by oxidation contain sites that allow abstraction of an ion or atom, and provide resonance stabilization of the resulting radical or ion. Some examples of readily oxidizable functional groups are branched aliphatic hydrocarbons, aromatic ring-containing polymers, allylic hydrocarbons, ether linkages, phenol, alcohols, aldehydes, and amines (Ratner, 2004). The mode of initiation may be either via host-generated molecular species, e.g., byproducts of phagocytosis (macrophage and polymorphonuclear leukocyte (PMN) metabolism), or by device or environmentally mediated species, e.g., metal ions. In either case, the result is often a severe decrease in molecular weight and tensile strength, and surface microcrack formation. Residual stresses generated during the manufacturing process and retained in the bulk may exacerbate surface degradation and enlarge cracks that ultimately lead to device failure. This phenomenon is referred to as environmental stress cracking (ESC). Polyether urethane (PEU), which is resistant to hydrolysis in vivo, has been used in numerous medical devices, including pacemaker leads. It was not until a decade after the first leads were implanted that a small percentage of lead

Degradation Products and Impurities

209

failures gave the first indications of PEU’s susceptibility to ESC (Santerre et al., 2005). Another degradation mechanism has been observed in PEU located in close proximity to corroding metal. The metal ions are strong oxidants that either directly, or via formation of intermediates, initiate chain scission. The solution for polyurethane biodegradation has been approached from multiple directions (Howard, 2002). Polyurethanes with enhanced oxidative stability have been developed by: ●

● ● ●

Substitution of the ether- and ester-soft segments with a variety of other monomers Forming a barrier by coating or surface functionalization Reduction of residual stresses by annealing Design of biodegradable polyurethanes (Santerre et al., 2005).

Calcification Calcification, the deposition of calcium-containing apatite-like mineral, affects a variety of synthetic and biologic biomaterials. It is the major cause of failure of bioprosthetic heart valves, and also affects synthetic PU valves, albeit less significantly. Other devices that have been reported to experience calcification include PU blood pump bladders, silicone breast implants, PET and PTFE vascular grafts, hydrogels (contact lenses and implants) (Ramani et al., 2005), urinary prostheses, and intrauterine contraceptive devices. Pathologic biomineralization may be regulated by the same physiological mechanisms that govern normal mineralization of collagenous tissue. While the pathophysiology is not completely understood, there are several common host and material factors that appear to play a role in calcification. Calcification is enhanced in the presence of necrotic cells and tissue, in regions undergoing frequent deformation (flexion, torsion), and in immature patients. In vitro studies have shown repeatedly that calcification can proceed in the absence of any cellular products. It has been proposed that the process occurs either via conformationally changed adsorbed protein molecules that bind calcium ions from surrounding media, or by absorption of calcium–protein complexes at the biomaterial/blood interface (Vasin et al., 1998). Hydrogel calcification has been shown to proceed by heterogeneous nucleation and spontaneous precipitation of calcium phosphate, with some dependence on polymer–free volume (Zainuddin et al., 2006). Zainuddin et al. studied hydrogel calcification by monitoring Ca+ ion diffusion through PHEMA membranes in a diffusion chamber. Chemical composition of the polymer appears to affect the initial calcification process, but does not play a role in later stages (Ramani et al., 2005). Animal implantation studies are commonly used for evaluation of biomaterials calcification. Large animal models have been used in the study of valve replacements, and small animals are used for subdermal implantation studies. Comparison of subdermal implants with explanted specimens indicates that subdermal calcification rates are significantly accelerated. Morphological characterization of the calcification of explanted specimens may be performed by visual or microscopic examination (light or SEM), energy dispersive X-ray spectroscopy (EDS), radiography, and histology. Chemical quantification of calcium and phosphorus deposits is obtained by atomic absorption and UV-Vis spectroscopy.

210

Safety Evaluation of Medical Devices

Assessment of calcification in vitro is often performed by immersion of the specimen in simulated body fluid (SBF) (Kokubo and Kushitani, 1990; Zainuddin et al., 2006). Kapolos et al. (1997) developed an in vitro model using a reactor containing supersaturated CaP solutions, under controlled pH and temperature. After immersion of the test specimen, the precipation process was monitored by a pair of glass-saturated calomel electrodes,. During the precipitation process, titrant was added to maintain a constant solution concentration, thereby obtaining an accurate measure of the rate of consumption. Mineral deposits formed in the reactor were morphologically similar to deposits found in explanted calcified implants, and were found to occur at similar sites (Mavrilas et al., 1999, 2004).

shown to cause a decrease in overall crystallinity and fatigue resistance. Annealing does not completely eliminate free radicals, and has been correlated with high levels of oxidation in accelerated aging studies. Radiation treated and subsequently annealed UHMWPE has been in clinical use since 1998, and have shown greatly reduced in vitro wear as compared with untreated materials. One of the most recent approaches to stabilization of radiation cross-linked treated UHMWPE incorporates the α-tocopherol form of vitamin E into the polymer, either in the resin prior to molding, or by diffusion into the molded component. Vitamin E is a known antioxidant that helps prevent oxidation by donating a hydrogen to the residual free radicals. It has been shown to significantly improve the oxidative stability of the cross-linked UHMWPE.

Wear Due to their relatively low modulus and inherent plasticity, polymers are particularly susceptible to wear, in varying amounts and degrees of severity. This is nowhere more evident than in total joint arthroplasty. In addition to hips and knees, there are now devices for replacement of shoulders, ankles, and finger joints, as well as vertebral disks. The most commonly accepted hip implants consist of a metal alloy articulating against UHMWPE. Since the 1990s, there have been numerous studies that confirmed a correlation between osteolysis and the degree of wear of the polyethylene bearing components of these implants (Dumbleton et al., 2002). It is generally agreed that macrophages play a key role in the osteolytic process. Polymer bearing surfaces are subjected to adhesion, abrasion, and fatigue in vivo. Visible or miscroscopic surface damage may occur in the form of scratching or pitting. Wear results in the loss of surface material and generation of debris. McKellop has proposed four modes of wear modes for bearing surfaces: 1. Mode 1: Two bearing surfaces contacting each other and moving under load as intended by designers. 2. Mode 2: Bearing surface wearing against non-bearing surface. 3. Mode 3: Primary bearing surfaces articulating with each other, but with third-body fragments interposed. 4. Mode 4: Two non-bearing surfaces moving against each other under load (McKellop, 2007). In vitro simulation of the adhesive and abrasive wear of UHMWPE bearing materials is typically performed using wear simulators. A pin-on-disc simulator loads UHMWPE pins on polished metal surfaces in a regular pattern for a predetermined number of cycles. A similar mechanism is the ring-on-disc simulator prescribed by ISO 6474. Whole joint simulators more accurately replicate the complex forces experienced by the components in vivo. In the simulator, components are surrounded by a synovial fluid analog, typically bovine serum. Unfortunately, due to the complicated relationships between materials properties and in vivo performance, in vitro wear tests are not always reliable predictors of clinical success. Researchers have used a variety of approaches to stabilize the polyethylene surface and reduce wear. Most of these involve cross-linking via ionizing radiation. The radiation must be followed by a stabilization step that eliminates residual free radicals that may result in oxidative degradation if left unmitigated. This has been accomplished through annealing or melting either during or immediately after the radiation treatment. Post-irradiation melting has been

Leaching and Migration Plasticizers Plasticizers are low molecular weight resins or liquids that improve the flexibility and processability of polymers by lowering the glass transition temperature (Tg). They are incorporated into the amorphous regions, where they interfere with secondary bonding between adjacent chains and increase free volume. External plasticizers are not attached to polymer chains, and can therefore be lost via evaporation, migration, or extraction.Migration may occur from solid to solid, solid to liquid (leaching), or solid to gaseous phase. Internal plasticizers are integrated into the polymer constitution, and therefore, remain part of the product. Thermal degradation is responsible for significant reductions in the effectiveness of both internal and external plasticizers (Rahman and Brazel, 2004). There has been a widely publicized debate regarding the potential carcinogenicity and endocrine modulating effects of phthalate plasticizers (Tickner et al., 2001). DEHP has been shown to cause liver toxicity and testicular atrophy in animals, but the potential effects on humans are still being investigated. PVC-based medical devices constitute 10% of the phthalate plasticizer market, and di(2-ethylhexyl) phthalate (DEHP) is found in high concentrations in blood bags (30–40% by weight) and dialysis tubing (up to 80%) (Tickner, 1999; Tullo, 2000). The detrimental dose of DEHP for adult humans has been estimated at 69 mg/kg per day. DEHP has relatively low aqueous solubility, but exhibits high lipid solubility that predisposes leaching into lipid rich liquids. Of particular concern are the high DEHP dose levels received by infants during blood transfusions, extracorporeal membrane oxygenation, and respiratory therapy. Adults and infants receiving intravenous total parenteral nutrition (TPN) through typical PVC tubing and connectors are also subject to high dose rates. Adult blood dialysis may also be a source of high DEHP doses. Concerns over phthalate exposure have prompted multiple scientific panel studies. In 2002, a study conducted by the National Toxicology Program (NTP) Center for the Evaluation of Risks to Human Reproduction (CERHR) led the FDA to suggest that manufacturers label certain devices with their DEHP content, or consider replacing PVC containing DEHP with alternative materials (Hileman, 2002). In 2005, continuing research and new concerns prompted NTP to conduct an updated evaluation of the potential for DEHP to cause adverse effects on the human reproductive system (Center for the Evaluation of Risks to Human Reproduction, 2006). The NTP expressed serious concerns that DEHP exposure to male fetuses or infants (via medical procedures or breastmilk)

Chapter 15:

Table 15.5

Degradation Products and Impurities

211

Reproductive Risks of Exposure to DEHP

Subjects

Estimated DEHP exposure level (µg/kg bodyweight/day)

Critically ill male infants Male infants younger than 1 year Male offspring of women undergoing certain medical treatments during pregnancy Male offspring during pregnancy Male children older than one year Adults

Concern level

6000 1–30* 1–30

Serious concern for adverse effects Concern for adverse effects Concern for adverse effects

1–30 1–30 1–30

Some concern for adverse effects Some concern for adverse effects Minimal concern for adverse effects

*Includes exposure during breast feeding. Source: Center for the Evaluation of Risks to Human Reproduction (2006).

may affect development of the male reproductive system. Other concerns are listed in Table 15.5. PVC is susceptible to thermal degradation and dehydrochlorination. Phthalate esters have been shown to reduce the rate of dehydrochlorination. As a result, PVC plasticizer migration results in discoloration, tackiness, and embrittlement (Rahman and Brazel, 2004). Several approaches to mitigation of plasticizer migration have been developed, with varying levels of success, including surface cross-linking, surface modification for improved hydrophilicity, impermeable surface coating, surface extraction, and addition of polymeric plasticizers (Rahman and Brazel, 2004). Additionally, there is intense ongoing research on development of alternative plasticizers for medical grade plastic. Biodegradable polymers present new challenges, since they will release additives during the resorption process.

Unreacted Monomer and Additives Incomplete polymerization or curing can result in leaching of unreacted monomer in solution. Extreme care must be taken in the design of in situ polymerizing implants, e.g., dental resins, adhesives, embolics, or bone fillers/cements. Dissolution of unreacted monomer has implications not only for toxicity, but also for decreasing the density and affecting the mechanical properties of the remaining polymer. Other typical polymer extractables include fillers, lubricants, antioxidants, pigments, slip agents, and products of side reactions formed during polymerization. Additionally, surface contamination may occur due to adsorption of volatile compounds, particulates, endotoxins, or even residues from packaging materials. Devices that are in contact with body fluids or tissues should be extracted in both water and non-polar solvents. The USP prescribes water and isopropanol extractions, and characterizes the total mass of nonvolatile residues without chemical characterization of the residues. ISO 10933-12 provides guidance for preparation of extracts of samples for biocompatibility assessment. Extraction can be used for one of three purposes: 1. Provide a test sample to be used in determining biological reactivity of leachables. 2. Identifying the potentially hazardous leachables. 3. Conducting human health risk assessment of the leachables. The purpose of the extraction, the nature and use of the final products, and the physicochemical properties of the device should be considered in selection of the extraction media and determination of the conditions of extraction. Extraction of temperature sensitive materials, e.g., polymers.,

should be conducted at temperatures below those that initiate material degradation. The duration of the extraction and the ratio of surface area to solvent should be adjusted to maximize the amount of material extracted and demonstrate the hazard potential of the device in humans. Polymer materials must be tested as small pieces to enhance immersion in the solvent. Elastomers, coated materials, laminates and composites are tested intact, since only the exposed surfaces are exposed to in vivo degradation. Extracts of materials that cure in situ should be collected at the point in the cure at which the materials are placed in situ. Both non-polar (e.g., water, PBS, culture media without serum) and polar (seed oil) solvents must be used. Additional media to be considered include diluted alcohols, DMSO, and culture media with serum. Volatile and non-volatile extractables are identified and quantified using sophisticated analytical instrumental methods, including: ● ● ● ● ●

Gas or liquid chromatography Infrared spectroscopy Mass spectroscopy Atomic absorption spectroscopy Inductively coupled plasma.

An example of an extractables study is included at the end of the chapter in the Appendix.

EFFECTS OF STERILIZATION Sterility Medical devices may be sold and used in two different states in terms of their bioburden (that is, the number of viable microorganisms present on the device)—either sterile or nonsterile. If they are intended to be used in the sterile state, the actual act of sterilization may be either performed by the manufacturer or by the user (though the latter is commonly the case only for reusable devices and instruments, such as surgical instruments). If the manufacturer is to sterilize the instruments, then this is specified in the DMF (and in product related literature), along with the means for achieving a suitable degree of sterilization and specification of a program for assuming performance of the process. ANSI/AAMI ST60 (1996) provides standards for indications of adequate sterilization of device products. FDA sets standards for actual sterilization results (FDA, 1982). The sterility assurance level (SAL) is the statistical probability of a device not being sterile after going through an established sterilization cycle. Residual microorganisms are usually measured as colony forming units (CFUs), with the amount of sterilization required to reduce the number of

212

Safety Evaluation of Medical Devices

CFUs by an order of magnitude for a specific device being termed the D10 (the dose to produce a tenfold reduction in bioburden). Doses here are both a concentration (or for radiation, a power level) and a time interval the device (or material) is exposed to that dose (Prince and Rubino, 1984; AAMI, 1997). There are four major means employed for sterilizing medical instruments, each of which has advantages and drawbacks.

Heating with Steam Heat (steam) is effective and inexpensive but can only be used for relatively small volume materials which are not degraded or deformed by it. Accordingly, it is typically useful only for smaller metal devices and cannot be used for materials such as plastics. It is most commonly used for reusable instruments and glassware. ISO 11134 sets forth requirements for the use of moist heat. Steam sterilization by autoclaving has traditionally been the most widely used method for medical instruments. Today, prevacuum, high-temperature steam sterilization is considered to be the safest and most practical means of sterilizing the majority of surgical instruments, surgical dressings, fluids, fabrics, and other absorbent materials. The process should be used with caution while sterilizing polymers and composites as both heat and steam can drastically alter the properties of these materials. During steam sterilization, polymeric materials are affected mainly by a deleterious effect that can be attributed to the hydrolysis of the polymer. This process leads to the creation of undesirable contaminants. Under prolonged steam autoclaving, 3–5 ppb of methylene dianiline has been detected (Mazzu and Smith, 1984) in the aqueous extract of methyl diisocyanate (MDI)-based polyurethane which was attributed to the hydrolysis of the polymer. Habermann and Waitzova (1985) observed the release of contaminating substances, whose structure was not elucidated, when polyvinyl chloride (PVC)-containing aqueous suspensions were autoclaved. Repeated autoclaving of PVC intended for biomedical applications was found to increase the mechanical properties of PVC for both covered and uncovered PVC samples. The change in the mechanical properties was attributed to the rearrangements in the PVC macromolecular chains giving rise to branches of varied lengths. Repeated autoclaving of uncovered PVC samples for 150 min was seen to induce increased leaching of the plasticizer. Seemingly stable polymers can also undergo changes in the surface morphology, especially if the glass transition temperature of the polymer is exceeded while autoclaving. This may cause changes in biocompatibility or blood contacting properties of the material. Oligomers or other inadvertent contaminants introduced subsequent to steam sterilization may have a significant effect on the biocompatibility and the performance of the polymers. A scanning electron microscopy study of different types of arterial prostheses made from Dacron® found that all the prostheses were coated with a layer of oligomers on autoclaving. The Cooley knitted Dacron grafts show the maximal amount of oligomer crystals, which was progressively increased with each subsequent successive resterilization. Additionally, this coating of oligomers resulted in increased hemolysis. Berger and Sauvage (1981) also noted late fiber deterioration in 493 Dacron arterial prostheses which were implanted for 3–15 years. One of the contributing factors to this deterioration was believed to be due to autoclaving.

Commercial poly (ethyleneterephthalate) (PET) is known to contain cyclic and linear oligomers as natural impurities, which are formed as a byproduct during polymerization and texturing. Cyclic trimer is a major oligomer component, and the content of cyclic trimer is actually a criterion for the quality of PET. Study of the effect of steam sterilization on PET has shown that oligomers and, in particular, the cyclic trimer increases in PET on repeated autoclaving. Studies (Nair, 1995) of the PET materials indicated increased values of heat of fusion and percent crystallinity, which suggested reorganization of the amorphous phase. Generation of new crystalline regions was ruled out as there was no increase in the intensity of the infrared peak of 973 cm−1 which indicates crystallinity. Chain scission of amorphous regions, resulting in increased incorporation of chain ends into existing crystallites, was, therefore, believed to be the reason for the increased values of crystallinity. The changes of the molecular weights (Mn and Mw) suggested that, though degradation would take place on autoclaving for 15 min, subsequent sterilization of 30 min results in solid-state polymerization or cyclization reactions. These cyclization reactions were responsible for the formation of fresh cyclic trimer molecules which could migrate to the surface. As moisture and air enhance oxidation and hydrolysis reactions, subsequent autoclaving for 60 min would result in more degradation of amorphous regions. The increased temperature could also facilitate the easy migration of the cyclic trimer to the surface.

Chemical Sterilants Chemical sterilants include ethylene oxide, glutaraldehyde, and formaldehyde. These can all be used for most devices that do not have any readily reactive (degradable) components. Ethylene oxide (ETO) is the most widely used, but is a mutagen, carcinogen, neurotoxin, and causes hypersensitivity responses in some individuals (Crammer et al, 1984; Chapman et al., 1986; Rockel et al., 1986). The concern with all of the chemical sterilants is residuals left on and permeated into devices. If any of the chemical sterilants are used, residual levels must be determined and kept below (or brought below by allowing the devices to “off gas”) certain established limits (HIMA, 1981). Table 15.6 states the ISO standards (ISO 11135) for allowable ETO residuals. Typically, ETO, epichlorohydrin (ECH) and ethylene glycol residual levels must all three be determined. Ethylene oxide is an effective bactericide active at temperatures as low as 60°C. As ethylene oxide can easily diffuse into materials being sterilized, it is possible to sterilize heat- or moisture-sensitive materials even through sealed plastic wrapping. One of the disadvantages of this method for polymeric materials is that they retain varying amounts of ethylene oxide. Residual ethylene oxide in sterilized plastic Table 15.6

ISO 10993–7: Ethylene Oxide Sterilization Residuals Average delivered dose of EO and ECH (mg/day)

Residue

Ethylene oxide Ethylene chlorohydrin

Permanent contact (>30 days) 0.1* 2

Prolonged exposure (1 to 30 days) 2 2

Limited exposure (<24 hour) 20** 12

Intraocular devices ≤ 0.5 µg/day. Extracorporeal blood purification set-ups ≤ 20 mg/d and 60 mg/month. Blood oxygenators and blood separators, ≤ 60 mg. Source: ISO (1993).

*

**

Chapter 15:

tubing has been reported to cause hemolysis of blood in heart–lung surgery. Adverse hemolytic reactions were also observed when blood was exposed to plastics that retained significant quantities of residual ethylene oxide. Ethylene oxide also possesses toxic properties and is a strong alkylating agent. The ethylene oxide molecule consists of two carbon atoms and an oxygen atom linked together in an unstable three-membered ring, allowing it to react with various functional groups, as shown in Figure 15.3. These are the sulfhydryl, amino, carboxyl, and hydroxyl groups of proteins and nucleic acids. The lethal effect of the ethylene oxide arises from its alkylating effect on these susceptible molecules. However, due to the reactivity of the gas with such functional groups, this type of sterilization should only be carried out at room temperature with materials such as polyurethanes, polyesters, and other polymers with potentially reactive groups. Several of the composite polymeric materials used currently, such as albuminated Dacron grafts or heparinized grafts, are liable to lose the beneficial effect extended by the albumin coating or heparin coating when subjected to ethylene oxide. Ethylene oxide sterilization of heparinized polymers reduces their nonthrombogenic properties (Bruck, 1971), while Guidoin et al. (1985) reported that ethylene oxide sterilized albuminated polyester grafts have a slower rate of healing when implanted in comparison to a similar graft sterilized by irradiation. Though retention of residual ethylene oxide is believed to be the cause for this slower rate of healing, it is likely that the alkylating action of ethylene oxide could denature the albuminated surface forming products that delay the healing process. Most manufacturers of ethylene oxide sterilizer equipment routinely recommend at least three and preferably five days aeration for ethylene oxide sterilized plastic tubings. However, the time for aeration is dependent on the material sterilized. Highly porous materials such as cellulose, paper, and natural rubber show very high values of ethylene oxide solubility and low diffusion coefficients. A study (Vink and Pleijsier, 1986) of residual ethylene oxide content in different polymers revealed that for a number of materials, the residual content was well above the levels which are currently considered to be safe even after aeration for fifteen days. For a reliable determination of aeration times after ethylene oxide sterilization of medical devices, the type of the material from which the device is made and, in particular, its thickness should be considered. Another point worth noting on the subject of adequate deaeration is the procedure for estimation of residual ethylene oxide. In a project involving the interlaboratory comparisons of the procedures for the estimation of ethylene oxide,

Ethylene Oxide

O R-C-OH

+

R-SH

HH R-S-C-C-OH HH

+

R-N-H2

HHH R-N-C-C-OH HH

O

CH2

Figure 15.3

CH2

O HH R-C-O-C-C-OH HH

+

Reactivity of ethylene oxide.

Degradation Products and Impurities

213

it was observed (Marlowe et al., 1987) that the average estimated total coefficient ranged from 8 to 22% even when following the same standard procedure of residual ethylene oxide estimation, emphasizing the need for standard and reliable procedures for estimating the residual ethylene oxide content.

Radiation Radiation provides the third means of sterilization. Gamma radiation or ultraviolet light are both used, with the former being predominant on the industrial scale. The advantages of radiation are that it can be used on sealed packages of material, leaves no residues, and can be used for large volumes of material. At the same time, there is a significant capital investment involved, safety precautions must be vigorously enforced, and some polymers are degraded (though this usually translates only to being discolored) by the absorbed energy in a gamma field. ISO 1137 sets requirements for radiation sterilization. Additionally, if a 60°C source is used, the sterilization time is long (AAMI, 1982). Biological indicator systems are typically used to ensure adequate sterilization performance (Prince, 1980). Irradiation of high molecular weight polymers results in either cross-linking or degradation, depending on the chemical nature of the system. Though the accepted dose for radiation sterilization is 2.5 Mrad, it is known that certain species require up to 6 Mrad for complete destruction. The effects of irradiation of increased doses or multiple sterilizations are assumed to be cumulative. Polymers which tend to cross-link show a steady increase in their molecular weights with increasing dose and with the formation of branches until a three-dimensional network is formed, while polymers that undergo degradation exhibit chain scission, the molecular weight steadily decreases with increasing radiation dose, and mechanical properties of the system simultaneously undergo changes. These properties include tensile strength, elastic modulus, shear strength, elongation, and occasionally, color. Though cross-linking and chain scission processes may occur simultaneously, usually one or the other predominates. In the case of vinyl polymers, if the carbon atoms of the main chain carry at least one hydrogen atom, the polymer tends to cross-link, whereas if the carbon atoms are fully substituted, the polymer tends to degrade. There is a correlation between cross-linking tendency and the heats of polymerization, showing that polymers which exhibit high heats of polymerization (above approximately 16 kcal/mole) cross-link, while polymers that exhibit low heats of polymerization (less than 16 kcal/mole) degrade. Irradiation of polymers can also result in gas evolution, double bond formation, and the production of trapped free radicals. The presence of impurities can accelerate the degradation process with the possibility of producing irritants or other undesirable products. Free radicals produced in polymers by irradiation in air convert to peroxidic radicals. Additionally, the irradiation process may result in discoloration of polymeric products. While general effects of irradiation on polymeric materials have been extensively studied (Chapiro, 1962), relatively small changes in physicochemical, mechanical, and biological properties may be tolerable in the short term, whereas similar changes may lead to catastrophic failures in long-term applications. This can result in seemingly unchanged materials which have different biological or mechanical properties because of minor undetectable damages. An example is the reported stability of cellulosic polymers up to 20 Mrad;

214

Safety Evaluation of Medical Devices

however, when the regenerated cellulose membranes are sterilized by radiation, Takesawa et al. (1987) have observed a reduction of vitamin B12 clearance, and an increase in hydraulic permeability of the membranes if the sterilization was carried out in the wet condition. Similarly, Rose et al. (1984) have observed that the wear of polyethylene exposed to γ irradiation increased with dosage and contact stress, becoming measurable in many cases only after a critical dose or stress was exceeded. The most significant effect noted was that the pressure dependence of the wear rate appeared to be a combination of chain scission and oxidation, suggesting that the radiation should be carried out in an inert atmosphere. Study of the effects of aging of γ radiation-sterilized isotactic polypropylene revealed that, though both crosslinking and degradation occur simultaneously during irradiation, branching takes place during post-irradiation aging. A dose of 2.5 Mrad was sufficient to introduce changes in crystallinity. These changes were attributed to the semicrystalline nature of the polymer and were observed to be different for the covered and uncovered samples. Additionally, the transitions between short- and long-range order and shortand long-duration stiffness in both covered and uncovered samples were affected by the formation of branches in the backbone of the polymer. A higher degree of branching was observed in covered samples, resulting in greater longduration stiffness in the transition between short- and longduration stiffness. The higher degree of branching in the covered samples was attributed to the high diffusion of energized oxygen into the polymer matrix when compared to the uncovered samples. PET, used extensively in biomedical applications, has been believed to withstand radiation sterilization without significant degradation. However, study of PET fibers revealed that a single dose of 2.5 Mrad was sufficient to induce chemical changes which were reflected in the changes of crystallinity and breaking load of the fibers. These changes were attributed to the degradation of PET in the amorphous regions and the recombination of degraded aliphatic segments. Baquey et al. (1987) and Takesawa et al. (1987) report the increased stability of Dacron or PET fabrics coated with albumin on irradiation, but the results of the investigation also indicate a leakage of albumin. The amount of albumin released was found to be dependent upon the nature of fabric used. Migration of chicken embryonic cells was also observed to be decreased for the irradiated fabrics. Another group of polymers used extensively in biomedical applications are the polyurethanes. Gamma irradiation of polyurethanes based on 4,4’ diphenyl methane diisocynate (MDI) have been reported (Shinatari and Nakamira, 1989) to produce the carcinogen 4,4’ methylene aniline with simultaneous detection of polyurethane oligomers. The γ irradiation of poly (DL-lactide/glycolide)-type microspheres has been reported to result in decreased molecular weights with the degradation continuing on storage. An inadvertent effect is the change in the release pattern of the drug loaded in the microspheres. Higher molecular weight polyglycolide sutures which were initially more resistant to enzymatic degradation become more prone to enzymatic attack as a consequence of the altered physical and chemical structure obtained on γ irradiation. Melberg et al. (1988) also report that the irradiation of PVC catheters of an external insulin pump produced chemical transformation products and damaged the insulin solution. Bacterial endotoxins were found to be unaffected by γ irradiation,

hence gloves and software sterilized by irradiation have been observed (Shumnes and Darby, 1984) to contain significant amounts of bacterial endotoxin. As another inadvertent effect, it was also observed that the endotoxin levels were increased when the bacterial counts were elevated, in some cases leading to contact dermatitis (Shumnes and Darby, 1984). The γ irradiation of intraocular lens, however, decreased the inflammatory reactions observed on implantation of ethylene oxide-sterilized lens. Ludwig et al. (1988) attributed this decreased response not only to less absorption of the toxic agent, but also to polymerization of any residual monomeric methylmethacrylate. Gamma radiation or ultraviolet light are both used, with the former being very predominant on the industrial scale. Radiation has significant advantages, but there is a significant capital investment involved, safety precautions must be vigorously enforced, and some polymers are degraded (though this usually translates only to being discolored) by the absorbed energy in a gamma field.

Dry Heat Dry heat is an additional available method, but one limited to use with metals and ceramics. It is the simple use of heat without any moisture present. There are also a number of new sterilization technologies available—variable pressure hydrogen peroxide, peracetic acid plasma, hydrogen peroxide plasma, ozone, intense pulsed light, and chlorine dioxide. The USP provides guidance on tests for verifying that devices have achieved the desired level of sterilization. These are sterility tests.

Sterility Tests The following procedures are applicable for determining whether a pharmacopoeial article purporting to be sterile complies with the requirements set forth in the individual monograph with respect to the test for sterility. In view of the possibility that positive results may be due to faulty aseptic techniques or environmental contamination in testing, provisions are included under Interpretation of Sterility Test Results for two stages of testing. Alternative procedures may be employed to demonstrate that an article is sterile, provided the results obtained are at least of equivalent reliability. Where a difference appears, or in the event of a dispute, when evidence of microbial contamination is obtained by the procedure given in this Pharmacopoeia, the result so obtained is conclusive of failure of the article to meet the requirements of the test. Similarly, failure to demonstrate microbial contamination by the procedure given in this Pharmacopoeia is evidence that the article meets the requirements of the test. The following considerations apply to sterilized devices manufactured in lots, each consisting of a number of units. Special considerations apply to sterile devices manufactured in small lots or in individual units where the self-destructive nature of the Sterility Test renders the conventional Sterility Test impracticable. For these articles, appropriate and acceptable modifications to the Sterility Test must be made. For articles of such size and shape as to permit complete immersion in not more than 1000 mL of culture medium, test the intact article, using the appropriate media, and incubate as directed under General Procedure. Proceed as directed under Liquids, beginning with “Examine the media visually.”

Chapter 15:

For devices having hollow tubes, such as transfusion or infusion assemblies, or where the size of an item makes immersion impracticable and where only the fluid pathway must be sterile, flush the lumen of each of 20 units with a sufficient quantity of fluid thioglycollate medium and the lumen of each of 20 units with a sufficient quantity of soybean-casein digest medium to yield a recovery of not less than 15 mL of each medium, and incubate with not less than 100 mL of each of the two media as directed under General Procedure. For devices in which the lumen is so small that fluid thioglycollate medium will not pass through, substitute alternative thioglycollate medium for fluid thioglycollate medium, but incubate the medium anaerobically. Where the entire intact article, because of its size and shape, cannot be tested for sterility by immersion in not more than 1000 mL of culture medium, expose that portion of the article most difficult to sterilize, and test that portion, or where practicable remove two or more portions each from the innermost portion of the article. Aseptically transfer these portions of the article to the specified number of vessels of appropriate media in a volume of not more than 1000 mL. and incubate as directed under General Procedure. Proceed as directed under Liquids, beginning with “Examine the media visually.” Where the presence of the test specimen in the medium interferes with the test because of bacteriostatic or fungistatic action, rinse the article thoroughly with a minimal amount of rinse fluid (see under Diluting and Rinsing Fluids.) Recover the rinse fluid, and test as directed for Devices under Test Procedures using Membrane Filtration (USP, 1996).

Degradation Products and Impurities

215

Interpretation Of Sterility Test Results First Stage At the prescribed intervals during and at the conclusion of the incubation period, examine the contents of all of the vessels for evidence of microbial growth, such as the development of turbidity and/or surface growth. If no growth is observed, the article tested meets the requirements of the test for sterility. If microbial growth is found, but a review in the sterility testing facility of the monitoring, materials used, testing procedure, and negative controls indicates that inadequate or faulty aseptic technique was used in the test itself, the First Stage is declared invalid and may be repeated. If microbial growth is observed but there is no evidence invalidating the First Stage of the test, proceed to the Second Stage.

Second Stage The minimum number of specimens selected is double the number tested in the First Stage. The minimum volumes tested from each specimen and the media and incubation periods are the same as those indicated for the First Stage. If no microbial growth is found, the article tested meets the requirements of the test for sterility. If, however, it can be demonstrated that the Second Stage was invalid because of faulty or inadequate aseptic technique in the performance of the test the Second Stage may be repeated. Note—Where sterility testing is used as part of an assessment of a production lot or batch or as one of the quality control criteria for release of such lot or batch, see USP Sterilization and Sterility Assurance of Compendial Articles (1211).

Sterile Empty or Prefilled Syringes Sterility testing of prefilled syringes is performed by employing the same techniques used in testing sterile products in vials or ampouls. The direct transfer technique may be employed if the Bacteriostasis and Fungistasis determination has indicated no adverse activity under the test conditions. Where appropriate, the membrane filtration procedure may be employed. For prefilled syringes containing a sterile needle, flush the contained produce through the lumen. For syringes packaged with a separate needle, aseptically attach the needle, and expel the product in to the appropriate media. Pay special attention toward demonstrating that the outside of the attached needle (that portion which will enter the patient’s tissues) is sterile. For empty sterile syringes, take up sterile medium or diluent into the barrel through the needle if attached, or if not attached, through a sterile needle attached for the purpose of the test, and express the contents into the appropriate media. Devices that are purported to contain sterile pathways may be tested for sterility by the membrane filtration technique as follows. Aseptically pass a sufficient volume of Fluid D through each of not less than 20 devices so that not less than 100 mL is recovered from each device. Collect the fluids in aseptic containers, and filter the entire volume collected through membrane filter funnel(s) as directed under Liquids Miscible with Aqueous Vehicles, beginning with “Aseptically remove the membrane(s).” Where the devices are large, and lot sizes are small, test an appropriate number of units as described for similar cases in the section, Sterilized Devices, under Test Procedures for Direct Transfer to Test Media.

HEAVY METALS Leachable heavy metals are a concern both in devices per se (for the possible potential hazard they present to patients) and in devices and their packaging (both of which end up in landfills) for their potential pollution of the environment, with subsequent potential harm to man and wildlife. Traditionally, the heavy metals are defined as all those metals with atomic weights greater than sodium (23) that form soaps on reaction with fatty acids. In practice, the metals of concern are lead, aluminum, selenium, nickel, cadmium, zinc, chromium, arsenic, mercury, tin, antimony, and iron. The USP presents specific methods in sections 〈231〉, 〈241〉, 〈251〉, 〈161〉, 〈291〉 and 〈211〉. Extractors are performed in purified water and great care must be taken to avoid contamination of test apparatus and material (due to the ubiquitous presence of the heavy metals—particularly lead—in the industrialized environment). Other pharmacopoeias also prescribe methods for analysis, but these are all severely limited (Anon, 1995). The present pharmacopoeial heavy metals test by sulfide precipitation is about 100 years old. Today, instrumental analytical techniques like atomic absorption spectrometry (AAS), X-ray fluorescence spectrometry (SRFS), and inductively coupled plasma-optical emission spectrometry (ICP-OES) are widely used in the pharmaceutical and food industry, as well as in other industrial branches, for the specific determination of metal traces. Scientific publications on heavy metals contamination and pollution are based exclusively on such modern instrumental techniques.

216

Safety Evaluation of Medical Devices

History and Objectives of Compendial Heavy Metal Tests

Scope and Limitations of the Present Pharmacopeial Heavy Metals Test

Over the hundreds of years of their history, pharmacopoeias have evolved from compendia on drug selection and manufacture (compounding) into compendia for device and drug testing. One of the first analytical tests introduced in the early phase of this development (at the turn of the century) was the test for heavy metals. USP VIII (1905) included the first general test for heavy metals, Time-Limit Test for Heavy Metals. The aim of this test was defined as follows: “This test is to be used to detect the presence of undesirable metallic impurities in official chemical substances or their solutions; these should not respond affirmatively within the stated time.” The test had two steps: (1) sulfide precipitation in a strongly acidic range, and (2) sulfide precipitation in an ammonia-alkaline medium. The metals listed as undesirable were antimony, arsenic, cadmium, copper, iron, lead, and zinc. A general test for the separate specific determination of arsenic had already appeared in USP VI (1893). In USP XII (1942), there was a changeover to determination in an acetic-acid medium. Simultaneously, a comparison solution for lead was also introduced, and it was the “darkness” of this solution that was to serve as a permissible limit. The Swiss and German pharmacopoeias (Ph. Helvetica, Deutsches Arzneibuch) underwent a very similar, but to some extent more delayed, development. In addition to testing for heavy metals in general, pharmacopoeias also require specific testing for a number of individual heavy metals, such as nickel (in polyols and hardened fats), iron (in diverse substances), and lead (in sugars). An analysis of the history of the pharmacopoeial heavy metals testing shows that the following objectives were pursued. In the 19th century and in the early part of the present century, a number of heavy-metal compounds considered to have medicinal value were commonly used in pharmaceutical products. The test was therefore originally very broad in scope (detection of all colored and dark sulfides precipitated in acidic and alkaline solutions) to prevent the use of mislabeled products or of products containing inadvertent admixtures of heavy metal compounds. The later restriction to dark sulfides precipitated from weakly acid solution, with lead as a comparison standard, and the additional specific tests for individual elements such as arsenic and iron imply a fundamental change of perspective. Clearly, the purpose now was to detect contamination caused by toxicologically significant heavy metals coming from manufacturing equipment and processes. The conditions of detection chosen show that the focus of interest was on lead and copper, two elements formerly widely used in factory equipment (e.g., in water pipes, in copper and brass kettles, and in the lead chamber process used in the manufacture of sulfuric acid, an essential basic chemical substance used in numerous synthetic processes. This historical review leads to the conclusion that the heavy metals test in its present form was clearly neither designed to be a universal test nor meant to be understood as one, and it clearly does not allow for such an interpretation.

The test in its present form—sulfide precipitation in a weakly acidic medium and comparison against a lead comparison solution at a concentration (usually) of 10 ppm—is theoretically suitable for the determination of bismuth, copper, gold, lead, mercury, ruthenium, silver(I), and tin(II). In practice, however, this method has several serious limitations (Gerney and Moine, 1989): ●

●

●

●

Elements like cadmium, antimony, and arsenic are not covered by this test because of the different colors of their respective sulfides (the test is suitable for black or dark brown sulfides only) are only partially covered in presence of very high concentrations without providing reliable information about the true amount of impurity present (source for wrong conclusions). Frequently, a substance must be ignited before it is tested for heavy metals. In most cases, this leads to a considerable loss of analyte, a loss that is matrix-dependent as shown in the average recovery rates found by Blake for hydroxypropyl methylcellulose (HPMC) (2). Important metals used in modern production equipment or as catalysts, such as iron, chromium, and nickel, are missed completely. It is virtually impossible to differentiate between highly toxic and less toxic metals. The test is nonselective and barely semiquantitative.

A unified and updated international approach is currently being pursued by ISO (Blake, 1989).

APPENDIX A: EXTRACTABLES STUDY Study Purpose The main purpose of this initial study is to determine the extractable amount of chemical compounds from an implantable sensor. This initial study is the first phase in a two phased approach with the final endpoint being a full characterization and quantification of any compounds of interest, which migrate into a simulated environment. In the first phase, the test system will be evaluated as a total system for analysis utilizing aggressive extraction conditions in both a polar and non-polar extracts, as recommended in the ISO 10993 series. The resulting extracts will be analyzed by a variety of analytical techniques, to identify and quantify materials that may have migrated from the product contact material into the solution of interest. Gas Chromatography/Mass Spectrometry (GC/MS) will be used to analyze the extract samples for Volatiles Organics Compounds (VOC), and Semi-Volatiles Organics Compounds (SVOC). Liquid Chromatography/Mass Spectrometry (LC/MS) will be used to analyze the extract samples for target Non-Volatiles Organics Compounds (NVOC). An LC chromatogram utilizing a UV detector will be obtained to determine the presence of peaks corresponding to non-target NVOC. Inductively Coupled Plasma (ICP) will be used to analyze the extract samples for metals.

Sample Preparation/Analysis The test article will be prepared in triplicate, utilizing extraction conditions more aggressive than typical use. This approach will aid in the determining what chemical compounds may be found to migrate into the contact fluid under typical use applications.

Chapter 15:

Extraction conditions and analysis are outlined in the Analytical Matrix depicted in Table 15.7. The following extraction conditions will be applied in this study: 1. Direct Component—No extraction: The test article will be placed in a headspace vial, heated, and analyzed for VOC by HEADSPACE GC/MS. An additional test article will be microwave digested and analyzed for metal content by ICP. Two test articles will also be analyzed by Fourier Transform Infrared to gain a material fingerprint. 2. Purified Water—Extraction conducted at 37°C for 72 hr, 4 weeks, and 13 weeks: The test article will be extracted in Purified Water utilizing a sample/solvent ratio recommended by ISO 10993-12 (or as specified by the sponsor) at 37 ± 2°C for 72 ± 2 hr, 4 weeks ± 4 hours, and 13 weeks ± 4 hours. The resulting extract will be analyzed following a purge-and-trap sampling for VOC. The nonvolatile residue of the original extract will be analyzed by FTIR. After a dichloromethane extraction of the original extract, and a subsequent concentration (if applicable), the resulting extract is analyzed to determine the SVOC by GC/MS, and the NVOC by both LC/MS and LC/UV. After a nitric acid digestion of the original extract, the digestate is subsequently analyzed to determine metal content by ICP. A blank purified water solution, utilizing the same sample preparation, but without any contact with the test article, will be prepared and analyzed. 3. 10% Ethanol in Water—Extraction conducted at 37°C for 72 hr, 4 weeks, and 13 weeks: The test article will be extracted in hexane utilizing a sample/solvent ratio recommended by ISO 10993-12 (or as specified by the sponsor) at 37 ± 2°C for 72 ± 2 hrs, 4 weeks ± 4 hours, and 13 weeks ± 4 hours. The resulting extracts will be analyzed following a purge-and-trap sampling for VOC. The nonvolatile residue of the original extract will be analyzed by FTIR. After a dichloromethane extraction of the original extract, and a subsequent concentration (if applicable), the resulting extract is analyzed to determine the SVOC by GC/MS, and the NVOC by both LC/MS and LC/UV. After a nitric acid digestion of the original extract, the digestate is subsequently analyzed to determine metal content by ICP. A blank 10% ethanol in water solution sample, utilizing the same sample preparation, but without any contact with the test article, will be prepared and analyzed.

Analytical Methods Headspace GC/MS—Volatile Organic Compounds Volatile organic molecules that may be released from the components arise from a variety of sources, such as monomers, residual solvents, residues from polymer treatments (e.g., washing), volatile degradation products and other volatile compounds. These compounds can be precisely and rapidly analyzed utilizing headspace gas chromatography with mass spectral detection. The test article material Table 15.7

Degradation Products and Impurities

217

can be analyzed without any extraction, by heating it in a closed vial and sampling the headspace volume for released compounds. Headspace GC/MS allows for the identification of the target compounds based on both the retention time of the analytes in the chromatogram and the mass spectrum of the eluting compound at this specific retention time. This instrumentation is also capable of scanning for a wide range of unknown analytes. The obtained mass spectra for the nontarget compounds are compared with electronic Mass Spectrum Libraries. This will assist in obtaining a qualitative and semi-quantitative identification of more than 170,000 non-target compounds. Additionally, a semi-quantitative analysis is also carried out via an internal calibration method to determine the concentration range of the compound in the polymer matrix. The analytical report will identify the volatile compounds, found in and on the polymer material, in the following manner: ●

●

●

●

IC: Identified Compound—The identity of the compound can be confirmed based upon the mass spectrum and retention time of the standard compound. MPC: Most Probable Compound—The identity of the compound is resolved by electronic Mass Spectral Libraries which gives an excellent fit (>80% fit) for the unknown compound. This compound, however, is not present as a standard to verify its retention time and mass spectrum, but in some cases may be acquired for confirmation as follow on work to the current study. TIC: Tentatively Identified Compound—The identity of the compound is determined as a best guess by electronic Mass Spectral Libraries which gives a lower fit (< 80% fit) for the unknown compound. If no Mass Spectral Library search is able to indicate any type of identification, the compound is reported as “unknown” and its mass spectrum may be provided at the sponsor’s request.

GC/MS—Volatile Organic Compounds The determination of monomers, residual solvents, volatile degradation products and other volatile compounds, present in the extracts, may also be precisely and rapidly analyzed following a purge-and-trap sampling by utilizing gas chromatography with mass spectral detection. In this sampling technique, the system is purged with helium, the gas is condensed in a carbon trap, and subsequently the trap is heated to release the volatile compounds which are injected in the GC/MS for separation and identification. GC/MS allows for the identification of the target compounds based on both the retention time of the analytes in the chromatogram and the mass spectrum of the eluting compound at this specific retention time. This instrumentation is also capable of scanning for a wide range of unknown analytes. The obtained mass spectra for the non-target compounds are compared with electronic Mass Spectrum Libraries. This will assist in

Analytical Matrix (At Each Time Point). Overview of the Sample Preparation and Respective Analyses, Per Technique

Extraction solvent

Extraction condition

ICP

GC/MS

GC/MS

LC/MS

LC/UV

FTIR

40°C for 24 hr

Direct component/ No extraction — × ×

Metals

Volatile compounds × × ×

Semi-volatile compounds × × —

Non-volatile compounds — × ×

Non-volatile compounds — × ×

Material Composition — × ×

None Purified Water 10% ethanol in water

× — —

218

Safety Evaluation of Medical Devices

obtaining a qualitative and semi-quantitative identification of more than 190,000 non-target compounds. The analytical report will identify the volatile compounds, found in and on the polymer material, in the following manner: ●

●

●

●

IC: Identified Compound —The identity of the compound can be confirmed based upon the mass spectrum and retention time of the standard compound. MPC: Most Probable Compound—The identity of the compound is resolved by electronic Mass Spectral Libraries which gives an excellent fit (>80% fit) for the unknown compound. This compound, however, is not present as a standard to verify its retention time and mass spectrum, but in some cases may be acquired for confirmation as follow on work to the current study. TIC: Tentatively Identified Compound—The identity of the compound is determined as a best guess by electronic Mass Spectral Libraries which gives a lower fit (< 80% fit) for the unknown compound. If no Mass Spectral Library search is able to indicate any type of identification, the compound is reported as “unknown” and its mass spectrum may be provided at the sponsor’s request.

GC/MS—Semi-volatile Organic Compounds Potential organic migration products may not be volatile enough to detect via purge-and-trap or headspace GC/MS. However, these thermostable compounds are still volatile enough to be studied via GC/MS (Gas Chromatography with Mass Spectral detection). These products are called the semi-volatile compounds. These compounds that may migrate into the solution originate from a variety of sources, such as: process lubricants, plasticizers, anti-oxidants, polymer degradation products, and solvents with a higher boiling point. The selected analytical method, a solvent extraction followed by injection into a GC/MS, allows for the identification of target analytes based on both the retention time of the analytes in the chromatogram and the mass spectrum of the eluting compound at this specific retention time. This instrumentation is also capable of scanning for a wide range of unknown analytes. The obtained mass spectra for the nontarget compounds are compared with electronic Mass Spectrum Libraries. This will assist in obtaining a qualitative and semi-quantitative identification of more than 190,000 non-target compounds. Additionally, a semi-quantitative analysis is also carried out via an internal calibration method to determine the concentration range of the compound in the polymer matrix. The Analytical report will identify the volatile compounds, found in and on the polymer material, in the following manner: ●

●

TIC: Tentatively Identified Compound—The identity of the compound is determined as a best guess by electronic Mass Spectral Libraries which gives a lower fit (< 80% fit) for the unknown compound. If no Mass Spectral Library search is able to indicate any type of identification, the compound is reported as “unknown” and it’s mass spectrum may be provided at the sponsor’s request.

ICP—Metal Content ICP analysis for Metal Content will be carried out for the direct component analyses.

Table 15.8

Target Metals—Standard List of Metals for Detection

Aluminum (Al) Antimony (Sb) Arsenic (As) Barium (Ba) Beryllium (Bc) Cadmium (Cd) Calcium (Ca) Chromium (Cr)

Cobalt (Co) Copper (Cu) Iron (Fe) Lead (Pb) Magnesium (Mg) Manganese (Mn) Nickel (Ni) Potassium (K)

Selenium (Se) Silver (Ag) Sodium (Na) Thallium (Th) Vanadium (V) Zinc (Zn)

Note:: Other metals can be added to the list, depending upon the available information of the composition of the materials.

A range of polymer additives, such as fillers (e.g., talc), pigments and catalyst residues are metal base complexes or do contain certain metals in their molecular structure. The presence of these polymer additives may then be seen via the presence of the corresponding metals. The test article material can be analyzed without any extraction, by direct microwave digestion of the test article. Material digestion reduces interference by organic matter and converts metals associated with particulates to a form that can be measured by inductively coupled plasma (ICP) spectroscopy. The presence of 22 standard metals can be determined via ICP analyses on the destruction solution. The detection limit for each metal will depend upon the amount of material used in the destruction. The detection limits typically range between 1–10 µg/L.

FTIR on NVR FTIR analysis will be carried out on the nonvolatile residue of the extracts. The presence of non-volatile residue in the extraction solution provides general information regarding the release of components. Residue on ignition provides general information related to type of components released from the material. Fourier Transform Infrared Spectroscopy (FTIR) is a technique used to analyze the chemical composition of many organic chemicals, polymers, adhesives, lubricants, coolants, gases, inorganics and minerals. An FTIR microscope accessory permits collection of spectra from a few nanograms of material. In some cases, thin films of residue may be identified.

Comprehensive Report The final report contains: ● ●

● ●

●

●

A detailed description of the test article. Description of analytical techniques carried out on the material, including sample preparation, extraction conditions, and instrumentation settings A list of all identified compounds, per analytical technique. For screening tests, the identification status of the reported compounds Quantitative and semi-quantitative (screenings) results of the compounds found in the test article. A full and detailed QA-review of all generated analytical data

The study cost as proposed is $73,750.00 for the total test system. This includes analytical testing, extraction, materials, study management, quality assurance, intermittent reporting after the 4-week time point evaluation is completed, final report, and archiving.

Chapter 15:

Remarks 1. This quotation does not include: ●

●

●

●

The determination of the asymptotic extraction behavior for the extraction solvents used. A comparative study between different extraction techniques. Contaminants study on the components of the assembled valve (to detect silicone oils, processing aids, etc.). Full validation of the analytical methods in the applied extraction solvents. It is evident that this can only be offered in detail once the initial results of the first controlled extraction studies are available.

2. The principles of SCT/AET, set forth in the PQRI document, can be applied in the study and can be included in the proposal. However, providing the correct numbers and calculations for these SCT/AET is the responsibility of the Sponsor. 3. If, based upon the interpretation of the analytical data, no identification of the extractable is possible, a further identification program can be offered via a combination of advanced MS and NMR techniques. 4. According to the FDA, extractables should be studied on all components, which can impact the “functionality” of the Products. The rationale behind this standpoint is that the Chemical composition is indicative for the physical performance of the component. The extractables profile is indicative for the Chemical composition. Therefore, the outcome of or the variation in extractables results may be indicative of the physical performance—and changes therein—of the components materials. In conclusion, the definition of “critical parts,” according to the FDA, is broader than the components, which have contact with the API or the patient. 5. Study outline in this quotation is for guidance and information purposes. Customized protocols can be provided for an additional cost upon receipt of detailed information on the test material and special test requirements as defined by the Sponsor. Should protocols be desired, protocol development will be completed prior to commencement of study. 6. Additional methods of analysis may be requested at additional cost and may impact project completion timeline. 7. All guidance documents, such as the FDA guidance document Container Closure Systems for Packaging Human Drugs and Biologics (1999), the EMEA guideline document Guideline on Plastic Immediate Packaging Materials (2005) and the PQRI document Safety Thresholds and Best Practices for Extractables and Leachables in Orally Inhaled and Nasal Drug Products, submitted by the PQRI Leachables and Extractables Working Group (September 2006) should be considered by the Sponsor for their appropriateness in their application. These guidance documents are only utilized to provide a general framework in the study design.

Quote Terms Payment is NET 30 days from the date of invoicing. Studies with a quotation of $10,000.00 or more will be invoiced 50% in advance, which is due prior to the commencement of the study. The remaining balance, and all additional charges, will be invoiced at study completion (issuance of draft and/or final report) under our normal 30-day terms. Full payment is

Degradation Products and Impurities

219

required prior to issuing the final report. Progress billing terms may also be applied to large projects and detailed in advance. THIS QUOTE IS VALID FOR 90 DAYS FROM THE DATE OF QUOTATION.

Prepared by Approval

Signature_________________________

Date___________

Terms and Conditions Additional costs Company X will perform the above named study in accordance with all applicable guidelines and a study protocol/outline. Additional fees will be incurred for all work or requests beyond the scope of the study, as stated in the protocol and/or guidelines. Additional testing, protocol amendments, reporting, special handling, retests and other requests, including consulting and custom protocol development, will be additionally charged. These fees include but are not limited to the following: Minimum Report/Study Fee—$50.00; Consulting—$175.00 to $300.00/hr (minimum 4 hours plus expenses); Revised Reports—$100.00/GLP, $50.00/NON-GLP; STAT Requests—Up to 100% of the test cost (minimum $200.00); Minimum Sample Return Fee— $100.00; On-Site Work—$2,500.00 (minimum plus expenses). Company X will assess a nominal charge of $1000.00 per day to our Sponsors for all directed regulatory inspections to support an application for market approval or data review. All Sponsors will be notified in advance of the inspection wherever possible

Study delay or cancellation Due to high costs involved with a study delay, including protocol development, technical scheduling, animal costs and rescheduling activities, Sponsor will be charged 10% of the study cost plus expenses when a study delay is attributable to the Sponsor. Study cancellation will result in the following charges: Up to 1 week before study initiation, per diem costs, expenses plus 10% of the study quotation; less than 1 week before study initiation, per diem, expenses plus 25% of the study quotation.

Reporting After completion of the study, one draft report may be issued. The Sponsor has up to 45 days in which to make comments regarding the draft report. Company X will work closely with the Sponsor to ensure the report is acceptable to both parties. After 45 days, if no comments or other remarks regarding the draft report have been made, Company X will issue a final report. A final report will only be modified by an agreed upon and signed Sponsor Communication Log (resulting in a study amendment). Interim reports will only be provided by Sponsor request and will be charged at $200.00 for NON-GLP studies and $1000.00 for GLP studies, if not agreed upon prior to study initiation. Sponsor specific amendments to a final report will incur charges of $100.00 for GLP and $50.00 for NON-GLP studies.

Study guarantee Company X conducts all studies with the utmost professional care and competence, but makes no qualifications or guarantees regarding study outcome. Company X does not

220

Safety Evaluation of Medical Devices

assume any liability for any actual or perceived failure of the study. Study materials may be returned to Sponsor as per instructions in the Test Requisition Form. Company X does not warranty or guarantee the condition of any material which has been utilized for testing purposes.

Sponsor Acceptance Protocols will be issued only upon acceptance of the quotation and receipt of the Sponsor P.O. Sponsor agrees to the terms as specified herein by signing below. Once signed, please return all of these contract documents to Company X along with your test sample(s) to: Attention: Sample Management Department. Sponsor Representative_______________________________ Date______________

REFERENCES AAMI (1982). Process Control Guidelines for Radiation Sterilization of Medical Devices, AAMI recommended practice, Association for the Advancement of Medical Instrumentation, Arlington, VA. AAMI (1997). ANSI/AAMI ST60—Sterilization of Health Care Products—Chemical Indicators. Part 1: General Requirements. AAMI. Anon (1995). Determination of Metal Traces—A Critical Review of the Pharmacopeial Heavy Metals Test. Pharmacopeial Forum, 21: 1638–1640. ASTM International (2006). ASTM F2129: Standard Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements to Determine the Corrosion Susceptibility of Small Implant Devices. Baquey, C., Sigot-Luizard, M.F., Friede, R.E., Proud’hom, R.E. and Guidoin, R.G. (1987). Radiosterilization of albuminated polyester prostheses. Biomat, 8: 185–189. Berger, K. and Sauvage, L.R. (1981). Late fiber deterioration in dacron arterial grafts. Ann. Surg., 193(4): 477–491. Black, D.L. (1991). 2, 4-TDA analysis with regards to PUF breast implants. Aegis Anal. Lab. Commun., 4: 1–2. Blake, K. (1989). Harmonization of the USP, EP, and JP heavy metals testing procedures. Pharm. Forum, 21(6). Bohner, M. (2000). Calcium orthophosphates in medicine: From ceramics to calcium phosphate cements. Injury, 31(4): 37–47. Brown, C. (2006). Biological effects of clinically relevant wear particles from metal-on-metal hip prostheses. Proceedings of the Institution of Mechanical Engineers. Part H, Journal of Engineering in Medicine, 220(2). Brown, C. (2007). Characterisation of wear particles produced by metal on metal and ceramic on metal hip prostheses under standard and microseparation simulation. Journal of Materials Science. Materials in Medicine, 18(5). Bruck, S.D. (1971). Sterilization problems of synthetic biocompatible materials. J. Biomed. Mater. Res., 5: 139–158. Bundy, K.J. (1994). Corrosion and other electrochemical aspects of biomaterials. Critical Reviews in Biomedical Engineering, 22(3-4): 139–251. Case, C.P., Langkamer, V.G., Howell, R.T., Webb, J., Standen, G., Palmer, M., Kemp, A. and Learmonth, I.D. (1996). Preliminary observations on possible premalignant changes in bone marrow adjacent to worn total hip arthroplasty implants. Clinical Orthopaedics and Related Research, 329 (Suppl): S269–79. Case, C.P., Langkamer, V.G., James, C., Palmer, M.R., Kemp, A.J., Heap, P.F. and Solomon. L. (1994). Widespread dissemination of metal debris from implants. Journal of Bone and Joint Surgery, British Volume, 76(5): 701–12. Catelas, I., Bobyn, J.D., Medley, J.J., Zukor, D.J., Petit, A. and Huk. O.L. (2001). Effects of digestion protocols on the isolation and characterization of metal-metal wear particles. II. Analysis of

ion release and particle composition. Journal of Biomedical Materials Research, 55(3): 330–337. Center for the Evaluation of Risks to Human Reproduction (2006). NTP brief on the potential human reproductive and developmental effects of di(2-ethylhexyl) phthalate (DEHP), draft. Chan, S.C., Birdsell, D.C. and Gradeen. C.Y., (1991). Detection of toluenediamines in the urine of a patient with polyurethanecovered breast implants. Clinical Chemistry, 37(5): 756–758. Chapiro, A. (1962). Radiation Chemistry of Polymeric Systems, WileyInterscience, New York. Chapman, J., Lee, W., Youkilis, E. and Martis, L. (1986). Animal model for ethylene oxide (ETO) associated hypersensitivity reactions. Trans. Am. Soc. Artif. Intern. Org., 32: 482–484. Chevalier, J. (2006). What future for zirconia as a biomaterial? Biomaterials, 27(4). Chevalier, J. (2007). Low-temperature degradation of zirconia and implications for biomedical implants. Annual Review of Materials Science, 37(1). Crammer, L.C., Roberts, M., Nicholis, A.J., Platts, M.M. and Patterson, R. (1984). IgE against ethylene oxide-altered human serum albumin in patients who have had acute dialysis reactions. J. Allergy Clin. Immunol., 74: 544–546. Doorn, P.F., Mirra, J.M., Campbell, P.A. and Amstutz, H.C. (1996). Tissue reaction to metal on metal total hip prostheses. Clinical Orthopaedics and Related Research, 329 (Suppl): S187–205. Doorn, P.F., Campbell, P.A., Worrall, J., Benya, P.D., McKellop, H.A. and Amstutz, H.C. (1998). Metal wear particle characterization from metal on metal total hip replacements: Transmission electron microscopy study of periprosthetic tissues and isolated particles. Journal of Biomedical Materials Research, 42(1): 103–111. Dumbleton, J.H., Manley, M.T. and Edidin, A.A. (2002). A literature review of the association between wear rate and osteolysis in total hip arthroplasty. The Journal of Arthroplasty, 17(5): 649–661. Ericsson, A.D. (1998). Syndromes associated with silicone breast implants: A clinical study and review. Journal of Nutritional and Environmental Medicine, 8(1): 35–51. FDA (1982). Sterilization of Medical Devices, Program 7378.008A. FDA Program Guidance Manual. FDA, Rockville, MD. Firkins, P.J., Tipper, J.L., Saadatzadeh, M.R., Ingham, E., Stone, M.H., Farrar, R. and Fisher, J. (2001). Quantitative analysis of wear and wear debris from metal-on-metal hip prostheses tested in a physiological hip joint simulator. Bio-Medical Materials and Engineering, 11(2): 143–157. Gerney, G. and Moine, J. (1989). Reflexions sur l’essai des metaux lourds. Pharmeuropa, 1(5). Gopferich, A. (1996). Mechanisms of polymer degradation and erosion. Biomaterials, 17(2): 103–114. Guidoin, R., Snyder, J., King, M. et al. (1985). A compound arterial prosthesis: The importance of the sterilization procedure on the healing and stability of albuminate dpolyester grafts. Biomat., 6(2): 122–128. Habermann, V. and Waitzova, D. (1985). On the early evaluation of extracts from synthetic polymers used in medicine. Arch. Toxicol, Suppl., 8: 458–460. Hallab, N., Merritt, K. and Jacobs, J.J. (2001). Metal sensitivity in patients with orthopaedic implants. The Journal of Bone and Joint Surgery, 83(3): 428. Heggers, J.P., Kossovsky, N., Parsons, R.W., Robson, M.C., Pelley, R.P., and Raine, T.J. (1983). Biocompatibility of silicone implants. Annals of Plastic Surgery, 11(1): 38–45. Heidemann, W., Fischer, J.H., Koebke, J., Bussmann, C. and Gerlach, K.L. (2003). In vivo study of degradation of poly-(D,L-) lactide and poly-(L-lactide-co-glycolide) osteosynthesis material. Mund-, Kiefer- Und Gesichtschirurgie: MKG 7(5): 283–288. Hileman, B. (2002). FDA suggests replacing DEHP in plastics. Chem Eng News, 80(37): 6. HIMA (1981). Guidelines for Evaluating the Safety of Materials Used in Medical Devices . . . Sensitization Studies, HIMA Document No. 10. Vol. 3, Washington, DC. (August). Howard, G.T. (2002). Biodegradation of polyurethane: A review. International Biodeterioration and Biodegradation, 49(4): 245–252. ISO (1993). ISO 10993-7: Ethylene Oxide Sterilization Residues.

Chapter 15:

ISO (1999). ISO 15814:1999—implants for surgery—copolymers and blends based on polylactide—in vitro degradation testing. Jacobs, J.J. (2006). Wear particles. Journal of Bone and Joint Surgery. American Volume, 88 (Suppl 2). Jacobs, J.J., Gilbert, J. and Urban, R. (1998). Current concepts reviewcorrosion of metal orthopaedic implants. Journal of Bone and Joint Surgery. American Volume, 80(2): 268–282. Kapolos, J., Mavrilas, D., Missirlis, Y. and Koutsoukos, P. G. (1997). Model experimental system for investigation of heart valve calcification in vitro. Journal of Biomedical Materials Research, 38(3): 183–190. Katti, K.S. (2004). Biomaterials in total joint replacement. Colloids and Surfaces.B, Biointerfaces, 39(3): 133–142. Kenny, S.M., and Buggy, M. (2003). Bone cements and fillers: A review. Journal of Materials Science. Materials in Medicine, 14(11): 923–938. Klein, C.P., Driessen, A.A., de Groot, K. and van den Hooff, A. (1983). Biodegradation behavior of various calcium phosphate materials in bone tissue. Journal of Biomedical Materials Research, 17(5): 769–784. Kokubo, T. and Kushitani, H. (1990). Solutions able to reproduce in vivo surface-structure changes in bioactive glass-ceramic A-W# 7# e. Journal of Biomedical Materials Research, 24(6): 721–734. Landes, C.A., Ballon, A. and Roth. C. (2006). In-patient versus in vitro degradation of P(L/DL)LA and PLGA. Journal of Biomedical Materials Research.Part B, Applied Biomaterials, 76(2): 403–411. Lide, D.R. CRC Handbook of Chemistry and Physics. 87th Ed, 2007. Lu, J., Descamps, M., Dejou, J., Koubi, G., Hardouin, P., Lemaitre, J. and Proust, J.P. (2002). The biodegradation mechanism of calcium phosphate biomaterials in bone. Journal of Biomedical Materials Research, 63(4): 408–412. Ludwig, K., Scheiffarth, O.F. and Von-Meyer, L. (1988). Reducing the amount of monomers in intraocular lenses through sterilization by gamma radiation, Ophthalmic. Res., 20(5): 304–307. Marlowe, D.E., Lao, NT., Eaton, A.R., Page, B.F. and Lao, C.S. (1987). Interlaboratory comparison of analytical methods for residual ethylene oxide in medical device materials, J. Pharm. Sci., 76(4): 333–337. Mavrilas, D., Apostolaki, A., Kapolos, J., Koutsoukos, P.G., Melachrinou, M., Zolota, V. and Dougenis, D. (1999). Development of bioprosthetic heart valve calcification in vitro and in animal models: Morphology and composition. Journal of Crystal Growth, 205(4): 554–562. Mavrilas, D., Kapolos, J., Koutsoukos, P.G. and Dougenis, D. (2004). Screening biomaterials with a new in vitro method for potential calcification: Porcine aortic valves and bovine pericardium. Journal of Materials Science. Materials in Medicine, 15(6): 699–704. Mazzu, A.L. and Smith, C.P. (1984). Degradation of extractable methylene dianiline in thermoplastic polyurethanes by HPLC. J. Biomed. Mater. Res., 18(8): 961–968. McKellop, H.A. (2007). The lexicon of polyethylene wear in artificial joints. Biomaterials, 28(34): 5049–5057. Melberg, S.G., Havelund, S., Villumsen, J. and Brange, J. (1988). Insulin compatibility with polymer materials used in external pump infusion systems. Diabetic Med., 5(3): 243–247. Moayed, M.H. and Newman, R.C. (2006). The relationship between pit chemistry and pit geometry near the critical pitting temperature. Journal of the Electrochemical Society, 153: B330. Nair, P.D. (1995). Currently practiced sterilization methods—Some inadvertent consequences. J. Biomat. Appl., 10: 121–135 Pihlajamaki, H., Salminen, S., Laitinen, O., Tynninen, O. and Bostman, O. (2006). Tissue response to polyglycolide, polydioxanone, polylevolactide, and metallic pins in cancellous bone: An experimental study on rabbits. Journal of Orthopaedic Research: Official Publication of the Orthopaedic Research Society, 24(8): 1597–1606. Prince, H.N. (1980). Characterization of Biological Indicators for Gamma Irradiation, In: Proceedings of the Third PMA Seminar on Validation of Sterile Manufacturing Processes: Biological Indicators. Pharmaceutical Manufacturers Association, Washington, DC.

Degradation Products and Impurities

221

Prince, H.N. and Rubino, J.R. (1984). Bioburden dynamics: the viability of microorganisms on devices before and after sterilization. Med Dev and Diagnos Industry, July. Rahman, M. and Brazel, C.S. (2004). The plasticizer market: An assessment of traditional plasticizers and research trends to meet new challenges. Progress in Polymer Science, 29(12): 1223–1248. Ramani, R., Parihar, J.K., Ranganathaiah, C., Awasthi, P., Alam, S. and Mathur. G.N. (2005). Free volume study on calcification process in an intraocular lens after cataract surgery. Journal of Biomedical Materials Research. Part B, Applied Biomaterials, 75(1): 221–227. Ratner, B.D. (2004). Biomaterials Science: An Introduction to Materials in Medicine. Academic Press. Rockel, A., Wahn, V., Hertel, J. and Fiegel, P. (1986). Ethylene oxide hypersensitivity reactions in dialysis patients. Lancet, 1: 382–283. Rose, R.M., Goldfarb, E.V., Ellis, E. and Crugnola, A.N. (1984). Radiation sterilization and the wear rate of polyethylene. J. Orthop. Res., 2: 393–400. Santerre, J.P., Woodhouse, K., Laroche, G. and Labow, R.S. (2005). Understanding the biodegradation of polyurethanes: From classical implants to tissue engineering materials. Biomaterials, 26(35): 7457–7470. Savarino, L., Granchi, D., Ciapetti, G., Stea, S., Donati, M.E., Zinghi, G., Fontanesi, G., Rotini, R. and Montanaro, L. (1999). Effects of metal ions on white blood cells of patients with failed total joint arthroplasties. Journal of Biomedical Materials Research, 47(4): 543–550. Shikanov, A., Kumar, N., and Domb. A.J. (2005). Biodegradable polymers: An update. Israel Journal of Chemistry, 45 (4): 393–399. Shintani, H. and Nakamura, A. (1989). Analysis of a carcinogen 4,4’ emthylene dianiline from thermosettins polyurethane during sterilization. J. Anal. Toxicol., 13: 354–357. Shumnes, E. and Darby, T. (1984). Contact dermatitis due to endotoxin in irradiated latex gloves. Contact. Dematitis., 10: 240–244. Singh, R. and Dahotre, N.B. (2007). Corrosion degradation and prevention by surface modification of biometallic materials. Journal of Materials Science. Materials in Medicine 18(5): 725–751. Sterilization of Medical Devices (1982), Program 7378.008A, FDA Compliance Program Guidance Manual, FDA, Rockville, MD, August. Syrett, BC. (1977). PPR curves—new method of assessing pitting corrosion resistance. Corrosion, 33(6): 221–224. Takesawa, S., Ohmi, S. and Konna, Y. (1987). Varying methods of sterilization and their effects on the structure and permeability of dialysis membranes. Nephral. Dial. Transplant, 1(4): 254–257. Tiainen, J., Soini, Y., Tormala, P., Waris, T. and Ashammakhi, N. (2004). Self-reinforced polylactide/polyglycolide 80/20 screws take more than 1(1/2) years to resorb in rabbit cranial bone. Journal of Biomedical Materials Research. Part B, Applied Biomaterials 70(1): 49–55. Tickner, J.A., Schettler, T., Guidotti, T., McCally, M. and Rossi, M. (2001). Health risks posed by use of di-2-ethylhexyl phthalate (DEHP) in PVC medical devices: A critical review. American Journal of Industrial Medicine, 39(1): 100–111. Tickner, J.(1999). The use of di-2-ethylhexyl phthalate in PVC medical devices: Exposure, toxicity, and alternatives. Available from http://www.noharm.org/library/docs/Use_of_Di-2-Ethylhexyl_ Phthalate_in_PVC_Med_3.pdf. Tullo, AH. (2000). Plastic additives steady evolution. Chemical Engineering News, 78(49): 1175–1179. Urban, R.M., Tomlinson, M.J., Hall, D.J. and Jacobs, J.J. (2004). Accumulation in liver and spleen of metal particles generated at nonbearing surfaces in hip arthroplasty. The Journal of Arthroplasty, 19(8, Supplement 1): 94–101. US FDA. Medical device recalls —recalls of zirconia ceramic femoral heads for hip implants. Available from http://www.fda.gov/ cdrh/recalls/zirconiahip.html. USP (1995). Heavy metals. United States Pharmacopoeia XXIII. U.S. Pharmacopelial Convention, Rockville, MD. USP (1996). Sterility Tests. United States Pharmacopoeia XXIII. U.S. Pharmacopelial Convention, Rockville, MD, pp. 1483–1488. Vasin, S.L., Rosanova, I.B. and Sevastianov, V.I. (1998). The role of proteins in the nucleation and formation of calcium-containing

222

Safety Evaluation of Medical Devices

deposits on biomaterial surfaces. Journal of Biomedical Materials Research, 39(3): 491–497. Viljanen, J., Pihlajamaki, H., Kinnunen, J., Bondestam, S. and Rokkanen, P. (2001). Journal of Orthopaedic Science: Official Journal of the Japanese Orthopaedic Association, 6(2): 160–166. Vink, P. and Pleijsier, K. (1986). Aeration of ethylene oxide-sterilized polymers. Biomaterials, 7: 30–48. Visuri, T., Pukkala, E., Paavolainen, P., Pulkkinen, P. and Riska. E.B. (1996). Cancer risk after metal on metal and polyethylene on metal total hip arthroplasty. Clin Orthop., 329: 280–289. von Burkersroda, F., Schedl, L., and Gopferich. A. (2002). Why degradable polymers undergo surface erosion or bulk erosion. Biomaterials, 23(21): 4221–4231. Weir, N.A., Buchanan, F.J., Orr, J.F., Farrar, D.F., and Dickson, G.R. (2004). Degradation of poly-L-lactide. Part 2: Increased temperature accelerated degradation. Proceedings of the Institution of Mechanical Engineers.Part H, Journal of Engineering in Medicine, 218(5): 321–330.

Wiltfang, J., Merten, H. A., Schultze-Mosgau, S., Schrell, U.,. Wenzel, D. and Kessler, P., (2000). Biodegradable miniplates (LactoSorb): Long-term results in infant minipigs and clinical results. The Journal of Craniofacial Surgery 11(3): 239–243; discussion 244–245. Yan, Y., Neville, A., Dowson D. (2006). Understanding the role of corrosion in the degradation of metal-on-metal implants. Proceedings of the Institution of Mechanical Engineers. Part H, Journal of Engineering in Medicine, 220(2): 173–180. Yoo, J.J., Kim, Y.M., Yoon, K.S., Koo, K.H., Kim, J.W., Nam, K.W. and Kim, H.J. (2006). Contemporary alumina-on-alumina total hip arthroplasty performed in patients younger than forty years: A 5year minimum follow-up study. Journal of Biomedical Materials Research. Part B, Applied Biomaterials, 78(1): 70–75. Zainuddin, D., Hill, J., Whittaker, A.K. and Chirila, T.V. (2006). In-vitro study of the spontaneous calcification of PHEMA-based hydrogels in simulated body fluid. Journal of Materials Science. Materials in Medicine, 17(12): 1245–1254.

16 Combination Products

Recent years have seen a vast increase in the number of new therapeutic products which are not purely drug, device, or biologic, but rather a combination of two or more of these. Classical examples are implanted drug delivery systems (whose primary function is drug delivery) and drug impregnated devices (in which drug delivery is an adjunct to the device function). Congress first acknowledged the need for specific regulation of such combination products in the 1990 Safe Medical Device Act.

FUTURE TRENDS Table 16.1 presents anticipated developments in the device combination product category between now and 2020 which lead to new clinical products. Three types of developments are generally expected. First, additional products designed for implanted delivery of insulin and other drugs. These include implanted pumps, possibly intelligent devices with improved biosensors to monitor concentrations in body fluids and make dynamic adjustments in delivery rates. Also there is the likely development of new polymeric timed-release devices which could improve the delivery of long-acting pharmaceuticals at optimized locations and rates. Second, new developments in drug-impregnated devices are expected. Examples included new types of cardiac implants with antithrombogenic and antiinfective drugs, as well as orthopedic implants with bacteriostatic coatings. Finally, under development are new developments in drug delivery systems to simplify reliable use by unsophisticated patients in home settings, including the increasing elderly population. Examples include nasal and inhalation products. Device regulation designation is by PMOA (principal mode of action), which is generally straightforward but can become less clear as precedents accumulate and technology becomes more complex. Although both extracorporeal and peritoneal dialysis systems are regulated as devices, dialysate concentrate for use with the former is a device but prepackaged dialysate for use with the latter is a drug. Sometimes consistency was elusive even when there was no combination, but just a single product. For example, in vitro diagnostics for detecting antibodies to HIV are regulated as biologics when they are used for screening the blood supply, but as medical devices when used for diagnostic or other screening purposes. When the FDA decides quickly and unequivocally on the regulatory status of a product, whether it was deemed a single product or was in combination with another product, there was relatively little opportunity for objection to the agency’s decisions about how to regulate combination products and products whose status was uncertain. In the case of blood devices, the EU has affirmed this process (Anon, 2000). In the Safe Medical Devices Act of 1990 (SMDA), Congress took these issues in hand and amended the Federal Food, Drug and Cosmetic Act (FDCA) to make it easier for the FDA to regulate combination products in a rational fashion. The new provisions altered the substantive provisions of the FDCA only in minor respects. The main thrust of

HISTORICAL BACKGROUND The history of this category includes a variety of producttypes, dating at least from the perfection of the hypodermic needle (1855). There are many modern examples of implanted delivery systems, such as the insulin pump (1980). One fundamental driving force for delivery systems has been the growth of new pharmaceutical products, especially since the dramatic expansion of drug research after 1945. That research has led to the synthesis and testing of millions of compounds for pharmacological and antimicrobial properties. Indeed, today much of that development is performed in automated computer-controlled systems, leading to an even greater acceleration of the process. The continued emergence of a stream of novel and more complex combination products has blurred any distinguishing lines of regulatory authority and has complicated product designation and regulation. The issue of products combing a device and a drug, such as an asthma inhaler, has received considerable scrutiny in recent years. But products combining a device and a biologic, such as organ replacement or assist devices, have received less attention. Recent trends, however, suggest that device and biologic combination products are quickly moving into the spotlight. A 1998 survey conducted by FDA identified hardware and tissue-engineered combination products as a rapidly growing trend in medical device technology (Herman et al., 1998). Even less than drug and device combinations, device and biologic products—which include, among other things, cellular and tissue implants, infused or encapsulated cells, artificial and replacement organs, heart valves and pumps, and cardiac, neural, and neuromuscular stimulation devices—do not fit neatly into existing regulatory paradigms. For example, as part of the question of regulation, FDA must take into account the possibility of tissue contamination and other hazards involved in using animal-derived tissues. What has resulted to date is a developing regulatory process. The written guidelines are fixed, but the day-to-day process is in flux (Merrill, 1994; March, 1998; Segal, 1999; Chapekar, 1996). 223

224

Safety Evaluation of Medical Devices

Table 16.1

Likelihood New Combination Device Drug Technologies

Biosensors Blood vessel prosthetics Bone prosthetics/growth Cardiac stimulation Cartilage prosthetics Computer aided clinical labs Drug impregnated devices Endoscopy Genetics—Cancer Hearing aids Heart pumps Heart valves Home diagnostics Image contrast agents Imaging: functional, content Implanted drug Integrated patient medical info systems Kidney prosthetics Laser surgery Liver prosthetics Min. invasive cardiology Min. invasive neurosurgery MRI Nanotechnology Nerve regeneration Neural stimulation Neuromuscular stimulation Ocular prosthetics Pancreas prosthetics Patient smart cards PET imaging Robotic surgery Skin prosthetics Telemedicine—Home use Telemedicine—radiology

Biosensors (for glucose, implantation, and systemic infection markers), Genetic diagnostics, Laser diagnosis & treatment, Minimally invasive devices Genetic therapy, Tissue engineered vessels, nerves, and devices Artificial organs, Tissue engineered devices Intelligent devices, Microminiaturized devices Tissue engineered device Computer aided diagnosis, Networks of devices Device/drug/biological products Minimally invasive devices, Telemedicine, Virtual reality diagnostics Genetic diagnostics, Genetic therapy Intelligent devices, Microminiaturized devices, Nonimplanted sensory aids Artificial organs Artificial organs, Tissue engineered devices, Device/drug/biological products Home/self monitoring & diagnosis Medical imaging Medical imaging, Minimally invasive devices, Networks of devices Biosensors, Device/drug/biological products, delivery systems, Home/self therapy, Intelligent devices, Robotic devices Computer aided diagnosis, Networks of devices, Telemedicine Artificial organs, Home/self therapy, Tissue engineered devices Laser diagnosis & treatment Artificial organs, Tissue engineered devices Minimally invasive devices vascular surgery Minimally invasive devices Greater resolution imaging Microminiaturized devices Tissue engineered devices Artificial organs, Electrical stimulation, Intelligent devices Electrical stimulation, Home/self therapy Artificial organs, Electrical stimulation, Intelligent devices Artificial organs, Tissue engineered devices Computer aided diagnosis, Networks of devices, Telemedicine Combined PET and CAT imaging Microminiaturized devices, Robotic devices Tissue engineered devices Home/self monitoring & amp; diagnosis, Telemedicine Telemedicine

the new law was managerial, directing the FDA to make decisions about which Center would have “primary jurisdiction” over a combination product, based on the agency’s understanding of the primary mode of action of the product. For these products, Center jurisdiction turns on the primary mode of action (PMOA). If the primary mode of action is that of a drug, then CDER has primary jurisdiction; if it is that of a device, jurisdiction is with CDRH; if that of a biological product, the Center for Biologics Evaluation and Research (CBER) has this jurisdiction. As the statute prescribed, the regulations go on to state that the Center with primary jurisdiction may consult with other agency components. Although neither the statute nor the regulations explain what “primary jurisdiction” means, it seems clear that the FDA intends it to mean that the Center that has primary jurisdiction will review the combination product and ordinarily give it just one approval, that is, an NDA, PMA, or biologic license application (BLA) as appropriate. Section 3.4(b) makes it clear, however, that the FDA’s designation of one agency component as having primary jurisdiction does not preclude, in appropriate cases, the requirement for separate application, e.g., a 510(k) and a BLA. When separate applications are required, both can be reviewed by the lead Center, but “exceptional” cases may involve a second application to be reviewed by a different Center. To facilitate this,

the agency published new delegations giving officials in each of the three Centers the authority to clear devices and to approve devices, drugs, biologics, or any combination of two or more of them (FDA, 1991). Contemporaneous with publication of the new regulations, the FDA made public three new Intercenter Agreements between CDRH and CBER, CDRH and CDER, and CDER and CBER. They describe the allocations of responsibility for numerous categories of specific products, both combination and noncombination. According to the regulations, these Intercenter agreements are not binding; they are intended to “provide useful guidance to the public” and, as a practical matter, to FDA staff as well. The Intercenter Agreements are a treasure trove of information. In addition to explicit guidance about which Center has the lead with respect to particular products and whether one Center or two will work on particular issues, they contain information and hints about whether the FDA believes it can regulate certain products at all, and if so, how (Pilot and Waldeermann, 1998; Adams et al., 1997). The regulations and Intercenter Agreements, however, do not answer every question, and the regulations recognize a role for the sponsor in cases of uncertainty. When the identity of the Center with primary jurisdiction is unclear or in dispute, or a sponsor believes its combination product is not covered by the Intercenter Agreements, a sponsor can

Chapter 16:

request a designation from the FDA’s product jurisdiction officer. A sponsor “should” file a request for designation with the product jurisdiction officer before submitting its application for marketing approval or an investigational notice. In practice, though, disputes or lack of clarity may not become evident until well into the review process, and it seems likely that the FDA would, if necessary, entertain requests for designation submitted at a later time. Section 3.7(c) of the regulations lists the information to be included in the request, all of which must fit on fifteen pages or less, including the identity of the sponsor, detailed information on the product, where the developmental work stands, the product’s known modes of action and its primary mode of action, and, importantly, the sponsor recommendation for which Center should have primary jurisdiction and the reasons for the recommendation. The FDA promises to check the request for designation for completeness within five working days of receipt, and to issue a letter of designation within sixty days of receipt of a complete request. If the FDA does not meet the sixty-day time limit, then the sponsor recommendation for the appropriate lead Center is honored. The agency’s letter of designation can be changed only with the sponsor’s written consent, or, if the sponsor does not consent, “to protect the public health or for other compelling reasons.” A sponsor must be given prior notice of any proposed nonconsensual change, and must be given an opportunity to object in writing and at a “timely” meeting with the product jurisdiction officer and appropriate Center officials. The Center for Devices and Radiological Health is designated the Center for major policy development and for the promulgation and interpretation of procedural regulations for medical devices under the Act. The Center for Devices and Radiological Health regulates all medical devices, inclusive of radiation-related devices, that are not assigned categorically or specifically to CDER. In addition, CDRH will independently administer the following activities (references to “Sections” are the provisions of the Act): 1. A. Small business assistance programs under Section 10 of the amendments (see PL 94-295). Both CDER and CDRH will identify any unique problems relating to medical device regulation for small business; B. Registration and listing under Section 510 including some CDER administered device applications. The Center for Drug Evaluation and Research will receive printouts and other assistance, as requested; C. Color additives under section 706, with review by CDER, as appropriate; D. Good Manufacturing Practices (GMPs) Advisory Committee. Under Section 520(f) (3), CDER will regularly receive notices of all meetings, with participation by CDER, as appropriate; and E. Medical Device Reporting. The manufacturers, distributors, importers, and users of all devices, including those regulated by CDER, shall report to CDRH under Section 519 of the Act as required. The Center for Devices and Radiological Health will provide monthly reports and special reports as needed to CDER for investigation and follow-up of those medical devices regulated by CDER. Table 16.2 presents the primary product responsibilities of CDER and CBER.

Table 16.2

Combination Products

225

Product Class Review Responsibilities

Center for Drug Evaluation and Review Natural products purified from plant or mineral sources Products produced from solid tissue sources (excluding procoagulants, venoms, blood products, etc.) Antibiotics, regardless of method of manufacture Certain substances produced by fermentation Disaccharidase inhibitors HMG-CoA inhibitors Synthetic chemicals Traditional chemical synthesis Synthesized mononuclear or polynuclear products including antisense chemicals Hormone products Center for Biologics Evaluation and Review Vaccines, regardless of manufacturing method In vivo diagnostic allergenic products Human blood products Protein, peptide, and/or carbohydrate products produced by cell culture (other than antibiotics and hormones) Immunoglobulin products Products containing intact cells or microorganisms Proteins secreted into fluids by transgenic animals Animal venoms Synthetic allergens Blood banking and infusion adjuncts

Device Programs that CDER and CBRH Each Will Administer Both CDER and CDRH will administer and, as appropriate, enforce the following activities for medical devices assigned to their respective Centers (References to “Sections” are the provisions of the Act): 1. A. Surveillance and compliance actions involving general controls violations, such as misbranded or adulterated devices under sections 301, 501, and 502; B. Warning letters, seizures, injunctions, and prosecutions under Sections 302, 303, and 304; C. Civil penalties under Section 303(f) and administrative restraint under Section 304(g); D. Non-regulatory activities, such as educational programs directed at users, participation in voluntary standards organizations, etc.; E. Promulgation of performance standards and applications of special controls under Section 514; F. Premarket Notification, Investigational Device exemptions including Humanitarian Exemptions, Premarket Approval, Product Development Protocols, Classification, Device Tracking, Petitions for Reclassification, post market surveillance under Sections 510(k), 513, 515, 519, 520(g) and (m), and 522, and the advisory committees necessary to support these activities; G. Banned devices under section 516; H. FDA-requested and firm-initiated recalls whether under section 518 or another authority and other Section 518 remedies such as recall orders; I. Exemptions, variances and applications of CGMP regulations under Section 520(f); J. Government-Wide Quality Assurance Program; and K. Requests for export approval under Sections 801(e) and 802.

226

Safety Evaluation of Medical Devices

Coordination The Centers will coordinate their activities in order to assure that manufacturers do not have to independently secure authorization to market their product from both Centers unless this requirement is specified in the Section VII.

Submissions Submissions should be made to the appropriate Center, as specified herein, at the addresses provided below: Address update: Food and Drug Administration Center for Drug Evaluation and Research (CDER) Central Document Room (CDR) 5901-B Ammendale Road Beltsville, Maryland 20705-1266 or Food and Drug Administration Center for Devices and Radiological Health Document Mail Center (HFZ-401) 9200 Corporate Blvd. Rockville, MD 20850 For submissions involving medical devices and/or drugs that are not clearly addressed in this agreement, sponsors are referred to the product jurisdiction regulations (21 CFR Part 3). These regulations have been promulgated to facilitate the determination of regulatory jurisdiction but do not exclude the possibility for a collaborative review between the Centers.

Center Jurisdiction The following subsections provide details concerning status, market approval authority, special label/regulatory considerations, investigational options, and inter-center consultations for the categories of products specified. Section VII provides the general criteria that CDRH and CDER will apply in reaching decisions as to which Center will regulate a product. A. 1. (a) Device with primary purpose of delivering or aiding in the delivery of a drug that is distributed without a drug (i.e., unfilled). EXAMPLES devices that calculate drug dosages drug delivery pump and/or catheter infusion pump for implantation iontophoreses device medical or surgical kit (e.g., tray) with reference in instructions for use with specific drug (e.g., local anesthetic) nebulizer small particle aerosol generator (SPAG) for administering drug to ventilated patient splitter block for mixing nitrous oxide and oxygen syringe; jet injector; storage and dispensing equipment. Status: Device and drug, as separate entities. Market Approval Authority: CDRH and CDER, respectively, unless the intended use of the two products, through labeling, creates a combination product.

Special Label/Regulatory Considerations: The following specific procedures will apply depending on the status of the drug delivery device and drugs that will be delivered with the device: (i) It may be determined during the design or conduct of clinical trials for a new drug that it is not possible to develop adequate performance specifications data on those characteristics of the device that are required for the safe and effective use of the drug. If this is the case, then drug labeling cannot be written to contain information that makes it possible for the user to substitute a generic, marketed device for the device used during developments to use with the marketed drug. In these situations, CDER will be the lead center for regulation of the device under the device authorities. (ii) For a device intended for use with a category of drugs that are on the market, CDRH will be the lead center for regulation for the device under the device authorities. The effects of the device use on drug stability must be addressed in the device submission, when relevant. An additional showing of clinical effectiveness of the drug when delivered by the specific device will generally not be required. The device and drug labeling must be mutually conforming with respect to indication, general mode of delivery (e.g., topical, IV), and drug dosage/ schedule equivalents. (iii) For a drug delivery device and drug that is developed for marketing to be used together as a system, a lead center will be designated to be the contact point with the manufacturer(s). If a drug has been developed and marketed and the development and studying of device technology predominates, the principle mode of action will be deemed to be that of the device, and CDRH would have the lead. If a device has been developed and marketed and the development and studying of drug predominates then, correspondingly, CDER would have the lead. If neither the drug nor the device is on the market, the lead center will be determined on a caseby-case basis. Investigation Options: IDE or IND, as appropriate. Inter-Center Consultation: CDER, when lead Center, will consult with CDRH if CDER determines that a specific device is required as part of the NDA process. CDRH as lead center will consult with CDER if the device is intended for use with a marketed drug and the device creates a significant change in the intended use, mode of delivery (e.g., topical, I.V.), or dose/schedule of the drug. A. 1. (b) Device with primary purpose of delivering or aiding in the delivery of a drug and distributed containing a drug (i.e., “pre-filled delivery system”). EXAMPLES nebulizer oxygen tank for therapy and OTC emergency use pre-filled syringe transdermal patch. Status: Combination product. Market Approval Authority: CDER using drug authorities and device authorities, as necessary. Special Label/Regulatory Considerations: None. Investigation Options: IND. Inter-Center Consultations: Optional.

Chapter 16:

A. 2. Device incorporating a drug component with the combination product having the primary intended purpose of fulfilling a device function. EXAMPLES bone cement containing antimicrobial agent cardiac pacemaker lead with steroid-coated tip condom, diaphragm, or cervical cap with contraceptive or antimicrobial agent (including virucidal) agent dental device with fluoride dental wood wedge with hemostatic agent percutaneous cuff (e.g., for a catheter or orthopedic pin) coated/impregnated with antimicrobial agent skin closure or bandage with antimicrobial agent surgical or barrier drape with antimicrobial agent tissue graft with antimicrobial or other drug agent urinary and vascular catheter coated/impregnated with antimicrobial agent wound dressing with antimicrobial agent. Status: Combination product. Market Approval Authority: CDRH using device authorities. Special Label/Regulatory Considerations: These products have a drug component that is present to augment the safety and/or efficacy of the device. Investigation Options: IDE. Inter-Center Consultation: Required if a drug or the chemical form of the drug has not been legally marketed in the United States as a human drug for the intended effect. A. 3. Drug incorporating a device component with the combination product having the primary intended purpose of fulfilling a drug function. EXAMPLES skin-prep pads with antimicrobial agent surgical scrub brush with antimicrobial agent. Status: Combination product. Market Approval Authority: CDER using drug authorities and, as necessary, device authorities. Special Label/Regulatory Considerations: Marketing of such a device requires a submission of an NDA with safety and efficacy data on the drug component or it meets monograph specifications as generally recognized as safe (GRAS) and generally recognized as effective (GRAE). Drug requirements, e.g., CGMPs, registration and listing, experience reporting, apply to products. Investigation Options: IND. Inter-Center Consultation: Optional. A. 4. (a) Device used in the production of a drug either to deliver directly to a patient or for the use in the producing medical facility (excluding use in a registered drug manufacturing facility). EXAMPLES oxygen concentrators (home or hospital) oxygen generator (chemical) ozone generator. Status: Device. Market Approval Authority: CDER, applying both drug and device authorities. Special Label/Regulatory Consideration: May also require an NDA if the drug produced is a new drug. Device requirements,

Combination Products

227

e.g., CGMPs, registration and listing, experience reporting will apply to products. Investigation Options: IDA or NDA, as appropriate. Inter-Center Consultation: Optional. A. 4. (b) Drug/device combination product intended to process a drug into a finished package form. EXAMPLES device that uses drug concentrates to prepare large volume parenterals oxygen concentrator (hospital) output used to fill oxygen tanks for use within that medical facility Status: Combination product. Market Approval Authority: CDER, applying both drug and device authorities. Special Label/Regulatory Considerations: Respective drug and device requirements, e.g., CGMPs, registration and listing, experience reporting will apply. Investigation Options: IDE or NDA, as appropriate Inter-Center Consultation: Optional, but will be routinely obtained. B. 1. Device used concomitantly with a drug to directly activate or to augment drug effectiveness. EXAMPLES biliary lithotriptor used in conjunction with dissolution agent cancer hyperthermia used in conjunction with chemotherapy current generator used in conjunction with an implanted silver electrode (drug) that produces silver ions for an antimicrobial purpose materials for blocking blood flow temporarily to restrict chemotherapy drug to the intended site of action UV and/or laser activation of oxsoralen for psoriasis or Cutaneous T-Cell Lymphoma Status: Device and drug, as separate entities. Market Approval Authority: CDRH and CDER, respectively. Special Label/Regulatory Considerations: The device and drug labeling must be mutually conforming with respect to indications, general mode of delivery (e.g., topical, IV), and drug dosage/schedule equivalence. A lead center will be designated to be the contact point with the manufacturer. If a drug has been developed and approved for another use and the development and studying of device technology predominates, then CDRH would have lead. If a device has been developed and marketed for another use and the development and studying of drug action predominates, then CDER would have lead. If neither the drug nor the device is on the market, the lead center will be determined on a case-by-case basis. If the labeling of the drug and device create a combination product, as defined in the combination product regulations, then the designation of the lead Center for both applications will be based upon a determination of the product’s primary mode of action. Investigation Options: IDE or IND, as appropriate. Inter-Center Consultations: Required. B. 2. Device kits labeled for use with drugs that include both device(s) and drug(s) as separate entities in one package with the overall primary intended purpose of the kit fulfilling a device function.

228

Safety Evaluation of Medical Devices

EXAMPLES medical or surgical kit (e.g., tray) with drug component. Status: Combination product. Market Approval Authority: CDRH, using device authorities, is responsible for the kit if the manufacturer is repackaging a market drug. Responsibility for overall packaging resides with CDRH. CDER will be consulted as necessary on the use of drug authorities for the repackaged drug component. Special Label/Regulatory Consideration: Device requirements, e.g., CGMPs, registration and listing, experience reporting apply to kits. Device manufacturers must assure that manufacturing steps do not adversely affect drug components of the kit. If the manufacturing steps do affect the marketed drug (e.g., the kit is sterilized by irradiation), and ANDA or NDA would also be required with CDRH as lead center. Investigation Options: IDA or IND, as appropriate. Inter-Center Consultation: Optional if ANDA or NDA not required. C. Liquids, gases or solids intended for use as devices (e.g., implanted, or components, parts, or accessories to devices). EXAMPLES dye for tissues used in conjunction with laser surgery, to enhance absorption of laser light in target tissue gas mixtures for pulmonary function testing devices gases used to provide “physical effects” hemo-dialysis fluids hemostatic devices and dressings injectable silicon, collagen, and Teflon liquids functioning through physical action applied to the body to cool or freeze tissues for therapeutic purposes liquids intended to inflate, flush, or moisten (lubricate) indwelling device (in or on the body) lubricants and lubricating jellies ophthalmic solutions for contact lenses organ/tissue transport and/or perfusion fluid with antimicrobial or other drug agent, i.e., preservation solutions powders for lubricating surgical gloves sodium hyaluronate or hyaluronic acid for use as a surgical aid solution for use with dental “chemical drill” spray on dressings not containing a drug component. Status: Device. Market Approval Authority: CDRH. Special Label/Regulatory Considerations: None. Investigation Options: IDE. Inter-Center Consultation: Required if the device has direct contact with the body and the drug or the chemical form of the drug has not been legally marketed as a human drug. D. Products regulated as drugs. EXAMPLES irrigation solutions purified water or saline in pre-filled nebulizers for use in inhalation therapy skin protectants (intended for use on intact skin) sun screens topical/internal analgesic-antipyretic.

Status: Drug. Market Approval Authority: CDER. Special Label/Regulatory Considerations: None. Investigation Options: IND. Inter-Center Consultations: Optional. E. Ad hoc jurisdictional decisions. EXAMPLES

motility marker constructed of radiopaque plastic brachytherapy capsules, needles, etc., that are radioactive and may be removed from the body after radiation therapy has been administered skin markers

STATUS

CENTER

Device

CDRH

Device

CDRH

Device

CDRH

Status: Device or drug. Market Approval Authority: CDRH or CDER as indicated. Special Label/Regulatory Considerations: None. Investigation Options: IDE or IND, as appropriate. Inter-Center Consultation: Required to assure agreement on drug/device status.

General Criteria Affecting Drug/Device Determination The following represent the general criteria that will apply in making device/drug determinations. A. Device criteria: 1. A liquid, powder, or other similar formulation intended only to serve as a component, part, or accessory to a device with a primary mode of action that is physical in nature will be regulated as a device by CDRH. 2. A product that has the physical attributes described in 201(h) (e.g., instrument, apparatus) of the Act and does not achieve its primary intended purpose through chemical action within or on the body, or by being metabolized, will be regulated as a device by CDRH. 3. The phrase “within or on the body” as used in 201(h) of the Act does not include extra corporeal systems or the solutions used in conjunction with such equipment. Such equipment and solutions will be regulated as devices by CDRH. 4. An implant, including an injectable material, placed in the body for primarily a structural purpose, even though such an implant may be absorbed or metabolized by the body after it has achieved its primary purpose will be regulated as a device by CDRH. 5. A device containing a drug substance as a component with the primary purpose of the combination being to fulfill a device function is a combination product and will be regulated as a device by CDRH. 6. A device (e.g., machine or equipment) marketed to the user, pharmacy, or licensed practitioner that produces a drug will be regulated as a device or combination product by CDER. This does not include equipment marketed to a registered drug manufacturer. 7. A device whose labeling or promotional materials make reference to a specific drug or generic class of drugs unless it is prefilled with a drug ordinarily remains a device regulated by CDRH. It may, however, also be subject to the combination products regulation.

Chapter 16:

B. Drug criteria: 1. A liquid, powder, tablet, or other similar formulation that achieves its primary intended purpose through chemical action within or on the body, or by being metabolized, unless it meets one of the specified device criteria, will be regulated as a drug by CDER. 2. A device that serves as a container for a drug, or a device that is a drug delivery system attached to the drug container where the drug is present in the container, is a combination product that will be regulated as a drug by CDER. Table 16.3

Combination Products

229

3. A device containing a drug substance as a component with the primary purpose of the combination product being to fulfill a drug purpose is a combination product and will be regulated as a drug by CDER. 4. A drug whose labeling or promotional materials makes reference to a specific device or generic class of devices ordinarily remains a drug regulated by CDER. It may, however, also be subject to the combination products regulation.

Examples of Existing Device/Drug Combination Products

Cardiac output catheter Extracorporeal sets Viscose/rayon dressings Cardiovascular oxygenator Paste bandages

Heparin Heparin Povidone iodine Heparin Clioquinol, coal tar, calamine

Medicated tulle dressings Antimicrobial drape Antiseptic wipes Cardiovascular guidewires Guidewires Antibiotic bone cement

Chlorhexidine Iodophore Chlorhexidine, centrimide, alcohol Heparin Heparin Antibiotic (for example, Gentamicin sulphate), colistin sulphomethate, sodium, erythromycin Heparin Heparin Silver Chlorhexidine digluconate Nonoxynol-9 Dexamethasone Mannitol

As device in UK As device in UK As drug in UK As drug in UK defoamer reservoir As drug in UK (if they have ichthammaol ancillary action) As drug in UK As device in UK As drug in UK As device in Spain As device in Spain, Switzerland, UK As drug (but soon to be regulated as device) As devices in Spain, Benelux, Italy As devices in Spain, Benelux, Italy As devices in three Benelux countries As device in Italy As device in Germany As device As device

Heparin Heparin Iodophore

Not applicable Not applicable Not applicable

Clioquionol (NaCa alignates, clauden powder)

As device in Germany

Collagen, albumen

As devices in UK

Extracorporeal cardiotomy reservoirs and filters Extracorporeal venous reservoirs and filters Bacteriostatic urological catheters Antiseptic island dressing Spermicidal condoms Pacemaker lead with a porous tip (seulte) Pacemaker lead with protector mannitol capsule (Sweet Tip) Biomedicus centrifugal pump Peripheral vascular cannulae Surgical gauzes or non-woven fabrics impregnated with iodophore Surgical gauzes or non-woven fabrics impregnated with alignates and Clioquinol Vascular prosthesis

REFERENCES Adams, D.G., Cooper, R.M. and Kahan, J.S. (1997). Fundamentals of Law and Regulation: An In-depth Look at Therapeutic Products. FDLI, Washington, DC. Anon (2000). EU Nations Agree that Directive Will Cover Human Blood Devices. European Drug and Device Report, 10(14): 1–21. CDER (1994). Premarket notifications: 510 (K) regulatory requirements for medical devices. HHS FDA, 1990: 4158. Chapekar, M.S. (1996). Regulatory Concerns in the Development of Biologic-Biomaterials Combination. J. Biomedical Materials Research, 33: 199–203. FDA (1991). Assignment of Agency Component for Review of Premarket Applications, Finale Rule, 56 Fed. Reg. 58,754 (Nov. 21, 1991).

Herman, W. A., Marlowe, D. F., and Harvey R. (1998) Future trends in Medical Device Technology, Biomedical Market Newsletter 8: 19–20. March, E. (1998). Combination Products: Who’s Regulating What and How—Part I, Pharmaceutical Engineering, July/August, pp. 34–38 Merrill, R. (1994) Regulation of Drugs and Devices: An evolution. Health Affairs, 13: 47–69. Pilot, L. and Waldemann, D. (1998). Food and Drug Modernization Act of 1997: Medical Device Provisions. Food and Drug Law Journal, 53: 267–295. Segal, S.A. (1999). Device and Biologic Combination Products— Understanding the Evolving Regulation. Medical Device and Diagnostic Industry, January, 1999, pp. 180–184.

17 Clinical Studies for Medical Devices

Clinical studies, once rare for devices other than Class III devices, are becoming much more frequently required and performed. Since 1991, ODE has taken actions for imposing more stringent requirements on clinical studies used to support device PMA applications. Clinical studies are also being required more often to support performance claims in 510(k) premarket notifications. The ODE focus on requiring carefully designed clinical trials is based, in part, on the Final Report of the Committee for Clinical Review, also known as the Temple Report. The Temple Committee reviewed a sample of PMA applications, IDEs, and 510(k) notices and found numerous deficiencies in the design, conduct, and analysis of clinical trials performed by sponsors in support of their applications. According to the Committee, the fundamental problem leading to inadequate data in most device applications was a lack of attention to basic study design. Specifically, the Committee’s report enumerated the following design deficiencies in the submissions examined:

In planning a device clinical trial, a sponsor must formulate a specific hypothesis to be tested (or question to be answered) in the study, and design a study that will answer the question. However, once the main hypothesis has been formulated, a number of other questions naturally arise: With what treatment (control) should results in the investigational device be compared? Is the objective of the study to show that the new device performs better or equivalently to an alternative therapy? How many patients (sample size) will be required to detect a clinically meaningful difference between the treatment and control groups with adequate statistical power, or ensure an acceptable degree of similarity? What is the diagnostic criterion by which to select study subjects? How will patients be assigned to the treatment and control groups? What baseline characteristics (characteristics of patients at the time of entry into the study) are important to prognosis? What will be the primary end point(s) of the study? What are the individual patient success/failure criteria? Will the patient, physician, or another be responsible for treatment administered? What follow-up information is important to collect? How will the safety of the device be established?

1. Failure to specify a clear hypothesis to be tested and to develop a clear plan to test it. 2. Failure to enroll a sufficient number of patients to answer the primary study questions. 3. Failure to adequately specify requirements for patients entering studies. 4. Failure to identify a control group. 5. Failure to properly assess the comparability of patients in the treatment and control groups. 6. Failure to clearly and precisely define end points. 7. Failure to implement blinded evaluation of end points, especially where end points are subjective in nature.

These questions, as well as others, must be addressed when planning a clinical study, rather than after the study has begun or the data collected. The methodology for responding to many of these questions is described in the following discussion of clinical study design parameters.

Recent findings support that FDA’s bioresearch monitoring program (BIMO) for devices continues to have problems (Anon, 2000). In this volume, of course, we are only concerned specifically with the aspects of clinical studies involved in evaluating device safety. General considerations of device clinical study design, execution, and analysis will be briefly considered to provide a context for this consideration, but those interested in broader issues should consult Spilker (1995) or Kahan (1994).

Study Hypothesis FDA’s ultimate regulatory question in evaluating whether to clear a device to market is whether it is safe and effective (PMA review), or whether it is substantially equivalent to a predicate product (510(k) review). These questions are, however, too nonspecific to be used to develop a clinical protocol intended to be pivotal in the regulatory clearance process for a device. The rationale for conducting a clinical trial is to determine whether an intervention, such as a medical device, has a particular postulated clinical effect. Thus, in designing a clinical trial, a clear statement regarding the objectives of the study should be made, and a specific hypothesis should be formulated indicating what proposition will be tested to determine if the investigational device is safe and effective. Consider, for example, a percutaneous transluminal coronary angioplasty (PTCA) device intended to open

DESIGN CONSIDERATIONS For any clinical study of a device conducted under an IDE, the sponsor is required to submit an investigational plan, including a protocol for the proposed study to ODE. The plan should describe three fundamental clinical trial areas: study design, study conduct, and data analysis. Each of these three elements needs to be carefully thought out in advance, long before the first patient is recruited into the study. 231

232

Safety Evaluation of Medical Devices

obstructed coronary vessels. The hypothesis could focus on whether the investigational device is as effective as a marketed device in achieving postoperative patency of all occluded vessels in a patient, or whether the investigational device is more effective than a predicate device in achieving both chronic patency (e.g., at six months post-procedure and acute patency. Thus, the first and key question for any study is the hypothesis to be tested. FDA and the device sponsor should always agree on the hypothesis before the first patient is enrolled. As noted in the Temple Report, a flawed hypothesis will lead to a flawed clinical trial where the data will not be sufficient to support a marketing application. In drafting the study hypothesis, the sponsor should always focus on the claims that will form the basis for marketing the device. Most marketing claims cannot be made for a Class III device without some foundation in the clinical study. Therefore, the study hypothesis and the design must be part of the company’s overall strategy in determining what marketing claims are necessary to have a commercially viable device.

Control Groups The Temple Report makes it clear that some type of control is essential to the conduct of a device clinical study. Having a control group permits the conclusion at the end of the study that observed differences in outcome and changes between the treatment group and control group are due to the device being studied and not to other factors. The Temple Committee noted that several device companies whose applications were reviewed failed to identify any control group when one was plainly needed to assess the safety and effectiveness of the device. Other companies, in the view of the Committee, failed to use the most appropriate control group or used poorly defined historical controls. For example, the Committee cited a submission for a defibrillator that was implanted in more than 300 patients, where results were reported without reference to any control group or any historical controls. While it appeared that survival for patients using the investigational defibrillator might have been similar to that reported in the literature for other devices, the agency had little information on which to make an assessment of efficacy. The literature on the clinical evaluation of medical devices is much smaller than that for most medicines. The quality of reported clinical trials varies enormously, and most trials suffer from lack of a control patient group that received a comparison medical device. Studies commonly describe a group of patients who received a device and who present before and after values for certain tests and parameters. There is usually no adequate control. One reason for this failure may be that the sponsors of most of these trials (i.e., device manufacturers) often do not wish to compare their device directly against a competitor’s. In addition it is rarely ethical to include a placebo or sham-operated group as a control. Placebo controls are often not necessary or possible in testing certain medical devices. Many situations in which Class II and III devices are used are not suitable for observation of a placebo effect (e.g., the patient may be unconscious, the interaction between patient and device may be minimal or nonexistent). There are several types of comparisons that may be used to evaluate medical devices (Table 17.1), and there are special considerations for designing protocols (Table 17.2) that illustrate differences between medical devices and medicines.

Table 17.1 Types of Comparisons that May Be Made in Trials of Medical Devices 1. 2. 3. 4. 5. 6.

Device A versus device B; usually performed single-blind Device A versus medicine B: usually performed single-blind Device A versus surgery B Device A versus no treatment Device A versus placebo device: preferably performed double-blind Device A versus other treatment modality (e.g., radiation)

The need of many patients for a Class III medical device is often certain (e.g., heart valve replacement, hip replacement). In such cases it is generally unethical to consider including a placebo treatment group in a clinical trial. Therefore, clinical trials depend heavily on the patient’s baseline values to evaluate the effects of the device. When open-label trials are conducted and no blinding is used, greater reliance should be placed on objective measures of change (e.g., ejection fraction) or semi-objective measures (e.g., distance the patient can walk without pain in a given period). Subjective parameters (e.g., quality of life measures) are usually more susceptible to influence by a placebo effect than are objective measures. Use of subjective measures should therefore be de-emphasized whenever possible, unless they are validated parameters or there is no other choice. When one product is compared with another the parameters evaluated must be comparable. Simply put, it will be very difficult in the near future to rely on historical controls for device clearance. If a sponsor feels historical controls are appropriate, detailed discussions with FDA are essential, especially concerning the issue of matching controls with active patients. If the control group will be one described in the literature, FDA and the sponsor must agree on whether the literature provides sufficient detail on the control patients to provide an adequate basis for comparison. In summary, reliance on historical controls at this point is discouraged by ODE and will require a clear explanation to ODE on why randomized concurrent controls are not feasible.

Eligibility Criteria Eligibility criteria characterize the study population and, thereby, impact on study design, ability to recruit patients, Table 17.2 Special Considerations in Designing Protocols for Medical Devicesa 1. In a double-blind trial it may be necessary to blind the identity of the device used; this may be done by disguising its outside appearance or by having the device inserted by one physician and having the patient evaluated by another physician. 2. Describe all increased risks that patients may experience in the trial; indicate how these risks will be minimized. 3. Label all devices appropriately, including the phrase, “CAUTION: INVESTIGATIONAL DEVICE LIMITED BY FEDERAL LAW TO INVESTIGATIONAL USE.” 4. Indicate appropriate methods of disposition of both used and unused devices. 5. Describe all functional tests of range of motion, correction of deformity, relief of symptoms, and other criteria of improvement. 6. Provide written instructions for use of the device, plus warnings, hazards, and contraindications.b a

These points are in addition to relevant considerations for designing protocols for medicine trials. b Evaluate the clarity of the instructions and their comprehension by potential users.

Chapter 17:

Clinical Studies for Medical Devices

233

and ability to generalize study results. Development of clear, unambiguous inclusion and exclusion criteria is essential in planning a clinical trial. It is necessary that the patients actually enrolled in the study form a subset of the general population defined by the eligibility criteria, and the results of a study can legitimately be generalized to other patients similar to those enrolled in the study. Thus, for example, if the only patients enrolled are adult males or postmenopausal females, it may be difficult to generalize any results to pediatric patients or young adult females. Formulation of eligibility criteria should be guided by a sponsor’s desire to demonstrate efficacy and its concern for patient safety. Patients who could benefit from the device are obvious candidates. Patients in whom hypothesized results of the device are likely be detected should also be studied. Any potential subject to whom the treatment is thought to be harmful or subjects likely to prematurely withdraw from the study should generally be excluded. To ensure comparability between groups, inclusion and exclusion criteria should be the same for the device and control groups. The group of enrolled patients can either be homogenous (patients with similar characteristics) or heterogeneous (where subjects differ in identifiable characteristics) in composition. A homogeneous study population may make the assessment of efficacy more straightforward, because the similarity of patients within the study population may not require analysis of study data by distinct prognostic subgroups. However, if eligibility criteria are too narrowly defined, recruitment may be hampered and study results may not be ready generalized. Heterogeneous populations may afford the opportunity to discover whether the device is effective in a separate subgroup of patients, but may necessitate a larger sample size if prognostically different subgroups are enrolled. A sponsor must carefully consider how homogeneous or heterogeneous the recruited patient population should be. This decision may largely rest on an assessment, based on preclinical information, of the kinds of patients for whom the device will likely be safe and effective. Typical inclusion and exclusion criteria relate to subject demographics such as age and sex, pregnancy status, projected lifespan, history of certain chronic diseases, use of concomitant medications, likelihood the patient will complete all follow-up, and presence of other confounding factors. Once an investigator has determined that a particular patient is eligible to participate in the study, a mechanism should be put in place to ensure that all eligible patients are offered the opportunity to participate in the study. FDA wants to ensure that investigators do not select patients subjectively, limiting the applicability of the device or masking some unidentified exclusion criteria. Marketing claims also should be considered in drafting eligibility criteria. If the company’s marketing experts say that a specific target patient population is necessary for ultimate commercial success, that population must be represented in the study.

may be disproportionately assigned to one group, thus obscuring the interpretation of any differences in outcome between the groups. It is generally accepted that randomization is the preferred method for reducing selection bias. Randomization tends to guard against imbalances of baseline prognostic factors between groups, protects against conscious or subconscious actions of study investigators that could lead to biased assignment of patients, and provides the probabilistic basis for most statistical analyses. Depending on the study, randomization procedures can be tailored to specific needs. For example, in block randomization, an equal number of patients is assigned to the various treatment groups from a specified number of enrollees (e.g., for every “block” of six patients randomizing three patients to the device group and three patients to the control group). Randomization can be carried out centrally by the sponsor or locally by each study investigator, especially if sizable variation is anticipated between centers. Stratification randomization (randomization of patients to study treatment groups within predefined strata) can also be used to achieve balance of patients within prognostic subsets of patients across the study groups.

Assignment of Intervention

Sample Size

Different methods have been developed for assigning patients to the investigational device group and the control group in a systematic manner that avoids selection bias. Selection bias occurs when, intentionally or unintentionally, patients with certain characteristics are more readily assigned to one treatment group than another. The result of selection bias is that patients who have important prognostic factors

Because the number of patients to be recruited into a study is critical, the issue of how many patients to enroll must be considered early in the planning stage. As observed by the Temple Committee, many clinical trials in which sample size requirements were not carefully considered lacked statistical power or the ability to detect device effects of clinical importance.

Double-Blind Trials There are a number of techniques whereby two or more medical devices may be effectively tested and compared in a double-blind manner. Once process is to have two (or more) surgeons each implant or insert different devices and to have an independent surgeon (who is kept blind) review all patients. Each surgeon may implant only one type of device, or may be permitted to implant all of the devices being compared. In the latter situation, devices should be implanted in a random order. Since many surgeons will have preferences as to the devices they use and also will be more experienced in implanting certain devices, they will generally prefer being assigned to only one type of device. Since differences in surgical skill rather than the device implanted may influence patient outcome, it is preferable to have several (or many) different surgeons each implant one specific device and a different group of surgeons each implant the other device studied. Alternatively, all surgeons may be trained in the device’s use so that they each achieve a minimal standard of technical competence. A third possibility is to measure and grade the sills of each surgeon and to consider this in the evaluation of the results. If multiple hospitals are involved in a trial of a medical device, it is often necessary to control for differences in hospital care. This may be done by having at least one surgeon at each hospital assigned to insert device number 1 and a different surgeon or group of surgeons at each hospital chosen to insert device number 2. It is also important to ensure that all patients meet standard criteria to enter the trial to eliminate the possibility that surgeons at one hospital select patients differently than surgeons at another hospital.

234

Safety Evaluation of Medical Devices

Computation of required sample size is based on testing the particular hypothesis stated by the sponsor. For example, the null hypothesis (the hypothesis to be tested) may be that the proportion of patients with a successful outcome in the investigational device group is the same as that in the control group. The alternative hypothesis may be that the proportion of successes is greater in the test device group. In testing the hypothesis, two types of errors can be made. A Type I error occurs when the null hypothesis is incorrectly rejected (i.e., concluding that the test device is better than the control, when in reality it is no better). A Type II error is made when the null hypothesis is incorrectly accepted (i.e., concluding no difference when the test device is better). The probability of a Type I error is referred to as alpha, and the probability of a Type II error as beta. The power of a statistical test (1 minus beta) is the probability that the null hypothesis is rejected if truly false; that is, power is the ability to detect a real difference, of a specified magnitude, between treatments. Typically, ODE prefers to see hypotheses tested at a 5% level of significance with at least 80 percent power. Factors that can affect determination of sample size include the type of primary end point analyzed, the desired size of the Type I and II errors, and assumptions about the anticipated success rates in the device and control groups. Planned subgroup analyses, anticipated significant betweencenter variation, prerandomization stratification, and anticipated dropouts may also affect the proposed sample size. Moreover, whether a study is designed to show a difference in effectiveness or equivalence to a predicate device will have a significant impact on sample size calculations. Intuitively, the larger the anticipated clinically meaningful difference between patients treated with a new device and patients treated with a control, the smaller the number of patients that will be required to demonstrate differences in the treatments. Similarly, the smaller the anticipated difference, the larger the number of patients that will be required to detect whether there is, in fact, a difference. Intuition may be less helpful, however, in determining an adequate number of patients when a company is trying to design a study to show its device is equivalent to another product. FDA typically requires that sponsors fully justify their proposed sample size by providing the following information: The hypothesis to be tested. The magnitude of a clinically meaningful difference between study groups. Assumptions about the anticipated performance of the device and control. The formula for computing sample size. The assumed size alpha and beta errors. Sample sizes for device clinical studies typically vary from less than 100 to nearly 300 patients for a therapeutic device. Several hundred to more than a thousand samples may be needed when studying a diagnostic device. The sample size for diagnostic studies is most often greater due to the need to establish the sensitivity and specificity of the diagnostic, which often cannot be done in an analysis of small numbers of samples that may or may not have the adequate number of positives to determine whether the diagnostic is efficacious.

End Points End points, or response variables, should be as clearly and precisely defined as possible. The sponsor should select outcome

variables that are the most informative, the most clinically relevant, and the least prone to bias. Stating the specific end points in advance facilitates planning of study design and calculation of sample size. Primary end points, which are designed to address the primary objectives of the study, form the principal basis for determining whether the device is safe and effective. Thus, the number of primary end points should be kept low to minimize confusion about the trial outcome. An individual patient’s outcome relative to a primary end point often results in the patient’s treatment being denied as a success or failure. Accordingly, the calculation of sample size is based on an analysis of the primary end points. Primary end points should be distinguished from secondary end points, which are designed to address secondary study objectives. End points can be objective or subjective, depending on the device and particular indications being studied, but should be capable of unbiased assessment. End points can be based on quantitative or categorical variables. Thus, an end point may show a change from one discrete state to another (e.g., living to dead), form one state to any of several others (e.g., from one disease stage to another), or from one level of a continuous variable to another (e.g., percent graft assimilation). Sometimes, response variables are combinations of variables, where events are combined if any one event occurs too infrequently to observe in a reasonable number of patients, or if a combination of responses is needed to determine if the patient has a successful outcome. A sponsor must also determine when, during the course of the study, the primary end point is to be measured (e.g., reduction in pain at six months after application of the device). Because of a history of prior misadventure, use of surrogate end points is often not well received by FDA. Surrogate end points are outcomes that are not themselves readily discernible as a clinical benefit to a patient, but which may be correlated with a clinical benefit. For example, an improvement in some hemodynamic parameter may show the device is working, but the improvement in this laboratory parameter may not be easily translated into a clinical benefit to the patient (i.e., mortality, morbidity). On the other hand, reduction in serum cholesterol may be an acceptable surrogate end point because its relation to a clinical benefit is well described in the scientific literature. However, where a surrogate end point is to be relied upon, some clear consensus from FDA that the end point is appropriate is essential. Examples of some broadly defined end points are 12-month patency rates for vascular grafts, 3-year rates of osseointegration for endosseous dental implants, and sensitivity and specificity of a device to diagnose cancer. The end points and success criteria should be discussed in detail with FDA prior to initiation of the study. Of course, it would be disastrous if the company were to choose end points, complete the study, and then have to start over again because FDA rejects the end points. Although FDA could reverse its judgment on the proper end points upon completion of the study, the sponsor should seek to reduce the risk of disagreement through careful planning and open and continuing communications with ODE.

Blinding (Masking) Blinding (recently re-titled masking) should be implemented to reduce several types of potential bias. If patients believe they are receiving a certain treatment, they may imagine certain beneficial or adverse effects resulting from the assigned treatment. If the investigators know the intervention

Chapter 17:

assigned, controls may not be followed as closely, or concomitant therapy may be disproportionately applied to one group of patients to compensate for the investigator’s potential bias. Bias can also occur when the person charged with evaluating response variable intentionally or inadvertently shades the outcome measures in favor of a particular treatment. Masking can take various forms in clinical trials. In an unblinded trial, both the patient and study investigator know which treatment has been assigned. In a typical singleblind study, only the investigator, not the patient, knows which treatment has been assigned. In a double-blind study, neither the subject nor the clinical investigator knows the assignment of intervention. Sometimes, modified double-blind studies are conducted, where the study investigator responsible for implementing the device knows the treatment assignment, but the observer responsible for evaluating safety and effectiveness outcomes is unaware of the treatment assignment. Although double-blind device studies are usually more complex and more difficult to conduct than other clinical trials, they are usually preferred by FDA. Triple-blind studies include those where, in addition to the patient and investigator being blinded, the committee monitoring or analyzing study results does not know the identity of the groups.

Comparability of Patients in Study Groups At baseline, before the investigational device is applied or the control is utilized, the control group should be similar in relevant respects to the device group, so that differences in outcome may be reasonably attributed to the action of the device being studied. If, for example, the treatment and control groups differ at baseline relative to factors that are known to affect outcome, it is often difficult to meaningfully compare the rates of therapeutic success in the two groups, even with statistical adjustment. Accordingly, it is essential that relevant factors be assessed at baseline to determine the comparability of the treatment and control groups, and to determine whether statistical adjustment is feasible if the groups are imbalances. As noted previously, one of FDA’s major concerns in utilizing historical controls is the fact that the control and device groups may not be comparable, and differences therefore may impact upon the safety and efficacy evaluation.

Other Design Considerations Decisions regarding the study design parameters described above have clear statistical implications. Other aspects of study design are equally important, but may involve less obvious statistical considerations. A sponsor must, for example, take the following actions when planning a study: It goes without saying that before any clinical study may be performed with a device, either the device must already be approved for such use in humans or an IDE (Investigational Device Exemption) must have been granted. There are very limited exceptions to this (primarily custom devices and veterinary devices). Consider the number of investigational sites to be employed (multicenter studies are essential, where the number of centers should be large enough to recruit the required number of patients, but small enough to pool results). Anticipate the duration of the study, including length of follow-up and recruitment period.

Clinical Studies for Medical Devices

235

Select study investigators who are sufficiently qualified to use the investigational device. Ensure informed consent is obtained from all patients. A sponsor should be prepared to specify each essential study design parameter before conducting a clinical trial of its device. This will demonstrate to CDRH that the sponsor knows the device and knows the expected results of the clinical trial, and, as a result, can confidently approach the agency. With today’s environment at FDA, the sponsor that is not sure which claims are appropriate or what end points to select should seriously consider conducting a feasibility study before entering into clinical trials.

Adverse Reactions When commonly expected types of complications from a medical device are known, the data collection forms may list these for the investigator to check off. Although this practice is not as common in clinical trials evaluating medicines, there are a number of advantages and justifications for its use in tallying “adverse reactions” due to medical devices. For example, the number of likely adverse reactions resulting form medical devices are usually fewer than for medicines and may be more precisely described. Federal regulations in the United States require manufacturers and importers of medical devices to report devicerelated serious injuries and deaths to the FDA. Physicians and other health professionals are encouraged to report their experiences to the manufacturer, directly to the FDA Product Monitoring Branch, or indirectly to the FDA via the Problem Reporting Program administered by the United States Pharmacopoeia. A newsletter of world medical device news (Clinica) is published by George Street Publications in Surrey, United Kingdom.

Study Contact Appropriate conduct of the clinical study is essential to ensure uniform implementation of the protocol and to obtain meaningful study results. During the planning phase, the likelihood of obtaining a sufficient number of patients in a specified period of time and the development of specific recruitment techniques must be evaluated. Data collection should be structured so that all data necessary to evaluate baseline comparability and outcome measures are collected. Efforts should also be directed at ensuring that key data are of high quality, and that missing and inconsistent information is minimized. Quality control methods should include techniques for data measurement, recording, transfer to computer databases, editing, verification, and changing of incorrect of missing data. Mechanisms should be put in place during the planning phase to ensure investigator compliance with the protocol and patient compliance with treatment and follow-up. Study monitoring, specifically required under the IDE regulations, is essential to ensure consistent implementation of the protocol and collection of quality data. Good clinical practice (GCP) regulations provide significant guidance on these issues, particularly the documentation aspects (FDA, 1981).

Study Data Analysis The Temple Report clearly recommended that the study protocol describe the methods of data analysis that are to be employed on the data collected in the study. The planned data analysis should include any plans to exclude certain treated patients from the analysis, any planned subgroup

236

Safety Evaluation of Medical Devices

comparisons, the identification of the primary end point from the several response variables, any planned interim analyses, and the specific statistical tests to be used. Because excluding randomized patients from analysis can lead to biased results, FDA prefers to see an intent-to-treat analysis, that is, an analysis of all randomized patients who received any treatment. Subgrouping on the basis of outcome variable scan also leads to biased results, and any planned subgroup analyses should, therefore, be set forth in the analysis plan. While it is desirable in a clinical trial to have groups of subjects that are comparable except of the treatment being studied, all prognostic factors may not be perfectly balanced. Covariate adjustment (a statistical method used to adjust outcome measures based on imbalances between groups at baseline) can be used to minimize the effect of such differences. Discussion of such adjustment and identification of potential covariates should be included in the analysis plan. A data analysis plan may also discuss contemplated pooling of results across various types of patients, for example, pooling across investigational sties, pooling across certain subgroups of patients, and potential pooling of U.S. and European clinical results. Where several end points will be analyzed to determine if the clinical study is a success, the issue of multiple comparisons also needs to be addressed in the data analysis plan. If several statistical tests are performed, it is likely that some will be significant by chance alone. Thus, if multiple end points are evaluated, statistical adjustment or increased sample size should be contemplated by the analysis plan. FDA prefers, in the interest of simplicity, that sponsors wait until all patients have been recruited before analyzing study data. There are situations, however, where sponsors may wish to analyze data in the middle of the clinical trial; for example, where there is a significant possibility of a safety problem or the potential for an overwhelming beneficial treatment response. Such analyses, termed “interim analyses,” are thus used, especially when a clinical study could be terminated early for safety or efficacy reasons. Since the use of interim analyses can affect the statistical interpretation of outcome measure, sponsors should discuss the possible use of these analyses in the IDE data analysis plan. One reason for FDA’s requirement of a prospective data analysis plan is to ensure that the sponsor is sufficiently confident in the design of the clinical trial to specifically identify how the device will perform. FDA frowns upon exploratory trials forming the pivotal basis for device approval, viewing such studies as “fishing expeditions” in which the study is conducted, and the collected data is analyzed, to find some result that can be presented as significant to the agency.

IRB Any studies conducted involving human subjects, no matter how trivial, must be performed with the review involvement, guidance, and approval of an institutional review board (IRB). The sponsor should have adequate nonclinical safety data in hand and be organized in such a manner as to cause the IRB to be comfortable that there is no unwarranted risk to patients in the trial being proposed. It is IRBs and not the FDA, which constitutes the primary gatekeeper for clinical research. Because IRBs play such a pivotal role in the review process for medical devices, it is important to understand the

requirements of IRBs. FDA has delineated the requirements for IRB membership, qualifications, functions, and operation in 21 CFR Part 56. An IRB is defined in the regulations as “any board, committee, or other group formally designated by an institution to review, to approve the initiation of, and to conduct periodic review of biomedical research involving human subjects” (21 CFR § 56.102(g)). The primary purpose of IRB review is to ensure the protection of the rights and welfare of the human research subjects. Highlights of the basic requirements for IRBs, as outlined in the regulations include: Membership. Each IRB must have at least five members, with varying backgrounds. Detailed requirements with regard to membership are established to prevent discrimination and protect against bias. Functions and operations. The IRB is required to follow written procedures with regard to conducting its initial review (discussed in Chapter 5) and continuing review of research. Written procedures are also required to ensure prompt reporting to the IRB and FDA of unanticipated problems that involve risks to human subjects or others. Informed consent. IRBs are required to ensure that information given to subjects as part of the informed consent is in accordance with the requirements for such consent set out in 21 CFR § 50.25. Continuing review. IRBs are required to conduct continuing review of research at intervals appropriate to the degree of risk, but not less than once per year. They have authority to observe, or have a third party observe, the consent process and the research. Suspension of approval. If research is not being conducted in accordance with the IRB’s requirements or if the study has been associated with unexpected serious harm to subjects, the IRB has the authority to suspend or terminate approval of the research. In this case, the IRB must report the suspension or termination to the investigator, institutional officials, and FDA. The report must include a statement of the reasons for the IRB’s action. FDA has increasingly focused its attention on the activities of IRBs in the past few years, demonstrating its authority over these bodies by stepping up its inspection of IRBs. Observations of noncompliance have resulted in numerous warning letters pertaining to deficiencies found in record keeping, protocol changes, and patient informed consent. If an IRB or the institution where a clinical investigation is being conducted fails to take adequate steps to correct deviations observed by FDA during such inspections, FDA may disqualify the IRB or parent institution if: The IRB has refused or repeatedly failed to comply with the application regulations. The noncompliance adversely affects the rights or welfare of the human subjects in a clinical investigation (21 CFR § 56.121(b)). FDA has told some IRBs to discontinue the approval of new studies until certain corrective actions have been taken. In fiscal year 1994, five IRBs were disqualified by FDA for failure to comply with appropriate regulations. After being disqualified, an IRB can become requalified upon the correction of all deficiencies; however, reinstatement requires a finding by the Commissioner that the IRB can meet all applicable standards in 21 CFR § 56.123.

Chapter 17:

Clinical Safety Studies There are some specific clinical studies which are performed with the single objective of ensuring device (or device material) safety. Most of these, such as those performed for contact lenses, vascular graft materials, and tampons, are specific for a device type. However, the repeat insult patch test (RIPT) is used only to evaluate safety and is common to all devices with significant human exposure. RIPT. The RIPT comes in several forms (Marzulli and Maibach, 1991), but as an example case the nine patch test is presented here. Induction phase. A sufficient amount of the test article (an amount to adequately cover the surface of the patch unit, approximately 0.1–0.15 g) was placed onto an occlusive patch, which was applied to the back of each subject between the scapulae and waist, adjacent to the spinal mid-line. The subjects were instructed to remove the patch 24 hr after application. Twenty-four hour rest periods followed the Tuesday and Thursday removals and 48-hr rest periods followed each Saturday removal. The sites were scored by a trained examiner just prior to the next patch application. This procedure was repeated every Monday, Wednesday, and Friday until nine applications of the test article were made. Procedurally, if a subject developed a positive reaction of a two-level erythema or greater during the induction phase or, at the discretion of the study director, if the skin response warranted a change in site, the patch would be applied to a previously unpatched, adjacent site for the next application. If a two-level reaction (or greater) occurred at the new site, no further applications would be made. However, any reactive subjects would be subsequently challenge patch tested. Challenge phase. After a rest period of approximately two weeks (no applications of the test article), the challenge patches are applied to previously unpatched (virgin) test sites. The sites are scored 24 and 72 hr after application. All subjects are instructed to report any delayed skin reactivity which might occur after the final challenge patch reading. When warranted, selected test subjects are called back to the investigators’ facility for additional examinations and scoring to determine possible increases or decreases in challenge patch reactivity. Skin responses for both the induction and challenge phases of the study are scored according to the following 6-point scale: 0 = No evidence of any effect + = Barely perceptible (minimal, faint, uniform, or spotty erythema) 1 = Mild (pink, uniform erythema covering most of the contact site) 2 = Moderate (pink-red erythema uniform in the entire contact site) 3 = Marked (bright red erythema with/without petechiae or papules) 4 = Severe (deep red erythema with/without vesiculation or weeping). All other observed dermal sequelae (i.e., edema, dryness, or hypo- or hyperpigmentation, etc.) are appropriately recorded and described as mild, moderate, or severe.

EPIDEMIOLOGY Epidemiology looks at the association between adverse effects seen in humans and a selected potential “cause” of

Clinical Studies for Medical Devices

237

interest, such as use of or exposure to a medical device, device material or contaminant. Those involved in the safety of medical devices must have a working knowledge of epidemiology, as it has been intimately involved in most of the major safety concerns (real and perceived) associated with medical devices during the last twenty years (as shall be reviewed in the last chapter of this book), and provides an essential tool for risk assessment (McMahon and Pugh, 1970; Gordis, 1988). Epidemiology is sometimes simply defined as the study of patterns of health in groups of people (Paddle, 1988). Behind this deceptively simple definition is an incredibly complex and diverse science, rich in concepts and methodology. The group of people of interest can be very small, historically consisting of as few as two people (Goudie et al., 1985) At the opposite extreme, studies of the geographic distribution of diseases using national mortality and cancer incidence rates have provided clues about the etiology of several diseases such as cardiovascular disease and stomach cancer. The patterns of health studied are also wide-ranging and may include the distribution, course, and spread of disease. The term “disease” also has a loose definition in the context of epidemiology and might include ill-defined conditions such as Ginger Jake and toxic shock syndromes, or consist of an indirect measure of impairment such as biochemical and hematological parameters of lunch function measurements. Epidemiology and toxicology differ in many other ways but principally in that epidemiology is essentially an observational science, in contrast to the generally experimental nature of toxicology. The opportunistic nature of epidemiology has been commented upon by several authors (e.g., Paddle, 1988; Utidjian, 1987). The epidemiologist often has to make do with historical data which have been collected for reasons which have nothing to do with epidemiology. Nevertheless, the availability of personnel records such as lists of new employees and terminations, payrolls and work rosters, and exposure monitoring data collected for compliance purposes, has enabled many epidemiological studies to be conducted in the occupational setting. Thus, the epidemiologist has no control over who is exposed to an agent or device, the levels at which they are exposed to the agent of interest, or to what other agents to which they may be exposed. The epidemiologist has great difficulty in ascertaining what exposure has taken place and certainly has no control over life-style variables such as diet, exercise, and smoking. Despite lack of precise data, the epidemiologist has one major advantage over the toxicologist: an epidemiology study documents the actual health experiences of human beings subjected to real-life exposures in an occupational, environmental, or clinical use setting. Indeed, Smith (1988) has recently expressed the view that the uncertainty in epidemiology studies resulting from exposure estimation may be no greater than or less than the uncertainty associated with extrapolation of results from animals to man. Regulatory bodies such as the U.S. Environmental Protection Agency (EPA) are starting to change their attitudes toward epidemiology and to recognize that it has a role to play in the process of risk assessment (Greenland, 1987). However, there is a continuing need for epidemiologists to introduce more rigor into the conduct of their studies and to introduce standards equivalent to the GLPs under which animal experiments are performed.

238

Safety Evaluation of Medical Devices

Measurement of Exposure Wegman and Eisen (1988) made the point that epidemiologists place much greater emphasis on the measure of response than on the measure of exposure. They claim that this is because most epidemiologists have been trained as physicians and are consequently more oriented toward measuring health outcomes. It is certainly true that a modern textbook of epidemiology such as that of Rothman (1986) provides minimal guidance about what the epidemiologist should do with exposure assessments. However, this is probably as much a reflection of the historical paucity of quantitative exposure information as a reflection on the background of epidemiologists. Nevertheless, it is surprising how many epidemiological studies do not contain even a basic qualitative assessment of exposure. The contrast between epidemiology and toxicology is never more marked than in the area of estimation of dose response. Not only can the toxicologist carefully control the conditions of the exposure to the agent of interest, but they can also generally be sure that his animals have not come into contact with any other toxic agents. The medical epidemiologist conducting a study of patients exposed to a hepatotoxin would certainly have to control for alcohol intake and possibly for exposure to other hepatotoxins in the clinical, work, and home environments. Nevertheless, it can be argued that epidemiology studies more accurately measure the effect on human health of real-life exposures. The epidemiologist must therefore frequently develop sampling strategies that generate exposure data suitable for both compliance and epidemiological purposes. For patients, of course, the situation is usually different. Quantitating exposures is generally straightforward If an exposure matrix has been constructed with quantitative estimates of the exposure in each device use and time-period, then it is a simple matter to estimate cumulative exposure. It is a more difficult process when, as is commonly the case occupationally, only a qualitative measure of exposure is available, e.g., high, medium or low. Even when exposure measurements are available, it may not be sensible to make an assumption that an exposure which occurred 20 years ago is equivalent to the same exposure yesterday. The use of average exposures may also be questionable, and peak exposures may be more relevant in the case of outcomes such as asthma and chronic bronchitis.

Epidemiological Study Designs What follows is a brief introduction to the most important types of studies conducted by epidemiologists, with an attempt to briefly describe the principles of the major types of epidemiological studies in order to assist the toxicologist in understanding the reporting of epidemiological studies and the assumptions made by epidemiologists.

Cohort Studies Historical Cohort Study. When the need arises to study the health status of a group of individuals, there is often a large body of historical data which can be utilized. If sufficient information exists on individuals exposed in past to a potential hazardous device material, then it may be possible to undertake a retrospective cohort study. The historical data will have been collected for reasons which have nothing to do with epidemiology. Nevertheless, the availability of medical records, and morbidity and mortality indices has enabled many epidemiological studies to be conducted (in particular, mortality studies).

The principles of a historical cohort study can also be applied to follow a cohort of patients prospectively. It should be emphasized that many historical data studies have a prospective element insofar as they are updated after a further period of follow-up. There is no reason why reproductive performance, incidence of serious bacterial infections, or almost any measure of health status of an individual, should not be studied retrospectively if sufficient information is available. Mortality and cancer incidence studies are unique among retrospective cohort studies in that they can be conducted using national cancer and mortality reports even if there has been no medical surveillance of the patient population of interest. A historical cohort study also has the advantages of being cheaper and providing estimates of the potential hazard much earlier than a prospective study. However, historical cohort studies are beset by a variety of problems. Principal among these is the problem of determining which patients have been exposed to the exact device of interest and, if so, to what degree. In addition, it may be difficult to decide what is an appropriate comparison group. It should also be borne in mind that in epidemiology, unlike animal experimentation, random allocation is not possible and there is no control over the factors which may distort the effects of the exposure of interest, such as smoking, alcohol use, and standard of living. Cohort Definition and Follow-Up Period. A variety of sources of information are used to identify patients exposed to a particular potential hazard, to construct a medical and occupational history, and to complete the collection of information necessary for follow-up. It is essential that the cohort be well defined and that criteria for eligibility be strictly followed. This requires that a clear statement be made about membership of the cohort so that it is easy to decide whether a patient is a member or not. It is also important that the follow-up period be carefully defined. For instance, it is readily apparent that the follow-up period should not start before exposure has occurred. Furthermore, it is uncommon for the health effect of interest to manifest itself immediately after the initiation of device use, and allowance for an appropriate biological induction (or latency) period may need to be made when interpreting the data. Comparison Subjects. The usual comparison group for many studies is the appropriate portion of the national population. However, it is know that there are marked regional differences in the mortality rates for many causes of death. Regional mortality rates exist in most industrialized countries, but have to be used with caution because they are based on small numbers of deaths and estimated population sizes. In some situations the local rates for certain causes may be highly influenced by the mortality of the patients being studied. Furthermore, it is not always easy to decide what the most appropriate regional rate of comparison purposes is, as many employees may reside in a different region from that in which the plant is situated. An alternative or additional approach is to establish a cohort of unexposed but otherwise comparable individuals for comparison purposes. For example, some studies of breast implant recipients have been restricted to female health care professionals and use others from this same group who have not received implants as a control group. However, patients with very low exposures to the device type of interest can often provide similar information. A good discussion of the issues is found in the proceedings of a conference entirely devoted to the subject (MRC, 1984).

Chapter 17:

Analysis and Interpretation. In a cohort study the first stage in the analysis consists of calculating the number of deaths expected during the follow-up period. In order to calculate the expected deaths for the cohort, the survival experience of the cohort is broken down into individual years of survival known as “person-years.” Each personyear is characterized by the age of the cohort member and the time-period when survival occurred and the sex of the cohort member. The person-years are then multiplied by age, sex, and time-period-specific mortality rates to obtain the expected number of deaths. The ratio between observed and expected deaths is expressed as a standardized mortality ratio (SMR) as follows: SMR = 100 ×

Observed deaths Expected deaths

Thus, an SMR of 125 represents an excess mortality of 25 percent. An SMR can be calculated for different causes of death and for subdivision of the person-years by factors such as level of exposure and time since first exposure. Interpretation of cohort studies is not always straightforward, and there are a number of selection efforts and biases that must be considered (Rothman, 1986). Occupational cohort studies routinely report that the mortality of active workers is less than that of the population as a whole. It is not an unexpected finding, since workers usually have to undergo some sort of selection process to become or remain workers. Nevertheless, this section effect, known as the “healthy worker” effect, can lead to considerable arguments over the interpretation of study results, particularly if the cancer mortality is as expected, but the all-cause mortality is much lower than expected. Similar possibilities must be considered for device recipient cohorts, who are likely to receive more medical services than the general population. Proportional Mortality Study. There are often situations where one has no accurate data on the composition of a cohort but does possess a set of death records (or cancer registrations). Under these circumstances a proportional mortality study may sometimes be substituted for a cohort study. In such a mortality study the proportions of deaths from a specific cause among the study deaths is compared with the proportion of deaths from that cause in a comparison population. The results of a proportional mortality study are expressed in an analogous way to those of the cohort study with follow-up. Corresponding to the observed deaths from a particular cause, it is possible to calculate an expected number of deaths based on mortality rates for that cause and all causes of death in a comparison group and the total number of deaths in the study. The ratio between observed and expected deaths from a certain cause is expressed as a proportional mortality ratio (PMR) as follows: PMR = 100 ×

Observed deaths Expected deaths

Thus, a PMR of 125 for a particular cause of death represents a 25% increase in the proportion of deaths due to that cause. A proportional mortality study has the advantage of avoiding the expensive and time-consuming establishment and tracing of cohort, but the disadvantage of little or no exposure information. Prospective Cohort Study. Prospective cohort studies are no different in principle from historical cohort studies in terms of scientific logic, the major differences being timing and methodology. The study starts with a group of apparently

Clinical Studies for Medical Devices

239

healthy individuals whose health and exposure is studied over a period of time. As it is possible to define in advance the information that is to be collected, prospective studies are theoretically more reliable than retrospective studies. However, long periods of observation may be required to obtain results. Prospective cohort studies or longitudinal studies of continually changing health parameters, such as lung function, incidence of inflammatory joint disease, blood biochemistry, and hematological measurements, pose different problems from those encountered in mortality and cancer incidence studies. The relationships between changes in the parameters of interest and device exposure measurements have to be estimated and, if necessary, a comparison made of changes in the parameters between groups. These relationships may be extremely complicated, compounded by factors such as aging, and difficult to estimate, as there may be relatively few measurement points. Furthermore, large errors of measurement in the variables may be present because of factors such as within-laboratory variation and temporal variation within individuals. Missing observations and withdrawals may also cause problems, particularly if they are dependent on the level and change of the parameter of interest. These problems may make it difficult to interpret and judge the validity of statistical conclusions. Nevertheless, prospective cohort studies provide the best means of measuring changes in health parameters and relating them to exposure.

Case-Control Study In a case-control study (also known as a case-referent study) two groups of individuals are selected for study, of which one has the disease whose causation is to be studied (the cases) and the other does not (the control). In the context of the chemical industry, the aim of a case-control study is to evaluate the relevance of past exposure to the development of a disease. This is done by obtaining an indirect estimate of the rate of occurrence of the disease in an exposed and unexposed group by comparing the frequency of exposure among cases and controls. Principal Features of Case-Control Studies. Case-control and cohort studies complement each other as types of epidemiological study (Schlesselman, 1982). In a case-control study the groups are defined on the basis of the presence or absence of a given disease and, hence, only one disease can be studied at a time. The case-control study compensates for this by providing information on a side range of exposures or other causes (background health problems of patients, for example) which may play a role in the development of the disease. In contrast, a cohort study generally focuses on a single exposure but can be analyzed for multiple disease outcomes. A case-control study is a better way of studying rare diseases because a very large cohort would be required to demonstrate an excess of rare disease. In contrast, a casecontrol study is an inefficient way of assessing the effect of an uncommon exposure, when it might be possible to conduct a cohort study of all those exposed. The complementary strengths and weaknesses of case-control and cohort studies can be used to advantage. Increasingly, mortality studies are being reported which utilize “nested” case-control studies to investigate the association between the exposures of interest and a cause of death for which an excess has been discovered. However, case-control studies have traditionally been held in low regard, largely because they are often badly conducted and interpreted. There is also a tendency to over interpret the

240

Safety Evaluation of Medical Devices

data and misuse statistical procedures. In addition, there is still considerable debate among leading epidemiologists themselves as to how controls should be selected, e.g., Poole (1986), and Schlesselman and Stadel (1987). Analysis and Interpretation. In a case-control study it is possible to compare the frequencies of exposures in the cases and controls. However, what one is really interested in is a comparison of the frequencies of the disease in the exposed and the unexposed. The latter comparison is usually expressed as a relative risk (RR), which is defined as: RR =

Rate of disease of exposed group Rate of disease of unexposed group

It is clearly not possible to calculate the RR directly in a case-control study, since exposed and unexposed groups have not been followed in order to determine the rates of occurrence of the disease in the two groups. Nevertheless, it is possible to calculate another statistic, the odds ratio (OR), which, if certain assumptions hold, is a good estimate of the RR. For cases and controls the exposure odds are simply the odds of being exposed, and the OR is defined as: OR =

Cases with exposure Controls with exposuree

Cases without exposure Controls without expposure

An OR of 1 indicates that the rate of disease is unaffected by the treatment being studied. An OR greater than 1 indicates an increase in the rate of disease in exposed workers. Matching. Matching is the selection of a comparison group that is, within stated limits, identical with the study groups with respect to one or more factors, such as age, years of device use or treatment, smoking history, etc., which may distort the effect of the exposure of interest. The matching may be done on an individual or group basis. Although matching may be used in all types of study, including follow-up and cross-sectional studies, it is more widely used in case-control studies. It is common to see case-control studies in which each case is matched to as many as three or four controls. Nested Case-Control Study. In a cohort study the assessment of exposure for all cohort members may be extremely time-consuming and demanding of resources. If an excess of death of incidence has been discovered for a small number of conditions, it may be much more efficient to conduct a casecontrol study to investigate the effect of exposure. Thus, instead of all members being studied, only the cases and a sample of noncases would be compared with regard to treatment history. Thus, there is no need to investigate the exposure histories of all those who are neither cases nor controls. However, the nesting is only effective if there are a reasonable number of cases and sufficient variation in the treatment of the cohort members.

morbidity statistics, in particular, those based on cancer registrations (Waterhouse et al., 1982). These health statistics can be used to study differences between geographic regions (e.g., maps of cancer mortality and incidence presented at a recent symposium; Boyle et al., 1978), device use, and time-periods. Investigations based on existing records of the distribution of disease and of possible causes are known as descriptive studies. It is sometimes possible to identify hazards associated with the development of rare conditions from observation of clustering in occupational groups, treatment groups, or geographical areas.

Cross-Sectional Studies Cross-sectional studies measure the cause (treatment and the effect (disease) at the same point in time. They compare the rates of diseases or symptoms of a treated group with an untreated group. Strictly speaking, the treatment information is ascertained simultaneously with the disease information. In practice, such studies are usually more meaningful from an etiological or causal point of view if the treatment assessment reflects treatment and past medical history. Current information is often all that is available but may still be meaningful because of the correlation between current treatment and relevant past treatments. Cross-sectional studies are widely used to study the health of groups of patients who are exposed to possible hazards but do not undergo complete regular surveillance (asthmatics and diabetics, for example). They are particularly suited to the study of subclinical parameters such as blood biochemistry and hematological values. Cross-sectional studies are also relatively straightforward to conduct in comparison with prospective cohort studies and are generally simpler to interpret.

Intervention Studies Not all epidemiology is observational, and experimental studies have a role to play in evaluating the efficiency of an intervention program to prevent disease (e.g., fluoridation of water). An intervention study at one extreme may closely resemble a clinical trial, with individuals randomly selected to receive some form of intervention (e.g., use of latex gloves). However, in some instances it may be a whole community that is selected to form the intervention group. The selection may or may not be random. The toxicologist might argue that even if selection were random, such a study of two communities, each consisting of many individuals, was is a sense a study of only two subjects. However, he should ask himself first whether the “three rats to a cage” design of many subacute toxicity studies really generates three independent responses per cage. Finally, the reliability of the collected data in epidemiology studies is, as with toxicology studies, reflective of the effort and care put into collecting the data. In particular, it should be kept in mind that data arising from medical assessments is much to be preferred to that generated by patient self reports.

Other Study Designs Descriptive Studies

CONCLUSION

There are large numbers of records in existence which document the health of various groups of people. Mortality statistics are available for many countries and even for certain devices and treatment types (e.g., Pell et al., 1978; Paddle, 1981). Similarly, there is a wide range of routine

There are fundamental differences between human and animal studies. Some of these are summarized in Table 17.3. But in the end, it is of course effects in humans that we are most concerned about. This gives the well-conducted and conclusive studies significantly greater weight than preclinical studies in

Chapter 17:

Table 17.3

Clinical Studies for Medical Devices

241

Differences Between Animal and Human Studies

Parameter

Animal study

Human study

Ethics

Provided that governmental animal cruelty/rights acts are not contravened, then it is perfectly acceptable to knowingly expose the animal to carcinogens, mutagens, teratogens, etc. Good laboratory practice (strict adherence to GLP) Monitored case histories (record of animal health throughout study) Regulated exposure (defined dose at defined intervals)

It is unethical to knowingly and deliberately expose humans to carcinogens, mutagens, teratogens, etc. Protocol for study (protocol may change during study) Exhaustive follow-up (sometimes subjects are untraceable/disappear) Defined exposed group (it may only be known whether there was a potential for exposure but not at what level) Various (depending on whether chemical is an occupational, marketplace, or environmental hazard. Mixed exposure at varied levels (usually “pulse” exposure) Valid exposed group (it can only be assumed that the only different variable is exposure) High degree of heterogeneity

Conduct Subject observation Dose

Length of exposure

Pattern of exposure Comparison groups Genetic homogeneity Death Relevance

Depending on the suspected effect of the chemical (generally lifetime for carcinogens, throughout organogenesis for teratogens, generations for reprotoxins) Single chemicals at around the maximal tolerated does, dose levels constant Randomized uniformity (control group known to have no exposure, otherwise identical with exposed group) Generally “inbred” strain used; hence, high degree of genetic homogeneity Standardized necropsy (every animal subject to pathological examination) Extremely relevant to the species in which data were generated (trans-species relevance unknown)

assessing human risk (Brown and Paddle, 1988; Glocklin, 1986; Freireich et al., 1966; Davidson et al., 1986; Calabrese, 1986).

REFERENCES Anon (2000). FDA’s BIMO Programs are Broken, OIG says. Guide to Medical Device Regulation Monthly Bulletin, August. Boyle, P., Muir, C.S. and Grundmann, E. (1989). Cancer Mapping. Springer-Verlag, Berlin. Brown, L.P. and Paddle, G.M. (1988). Risk assessment: Animal or human model? Pharm. Med., 3: 361–374. Calabrese, E.J. (1986). Animal extrapolation and the challenge of human heterogeneity. J. Pharm Sci., 75: 1041–1046. Davidson, I.W.F., Parker, J.C. and Beliles, R.P. (1986). Biological basis for extrapolation across mammalian species. Regulatory Toxicol. Pharmacol., 6: 211–237. FDA (1981). Protection of Human Subjects/Informed Consent/ Standards for Institutional Review Boards for Clinical Investigations. Federal Register, January 27. Freireich, E.J., Gehan, E.A., Rall, D.P., Schmidt, L.H. and Skipper, H.E. (1966). Quantitative comparison of toxicity of anticancer agents in the mouse, rat, hamster, dog, monkey, and man. Cancer Chemother. Rep., 80: 219–244. Glocklin, V.C. (1987). Current FDA perspective on animal selection and extrapolation. In: M.V. Roloff (Ed.), Human Risk Assessment: The Role of Animal Selection and Extrapolation. Taylor and Francis, London, pp. 15–22. Gordis, L. (Ed.) (1988). Epidemiology and Health Risk Assessment. Oxford University Press, New York. Goudie, R.B., Jack, A.S. and Goudie, B.M. (1985). Genetic and developmental aspects of pathological pigmentation patterns. Curr. Top. Pathol., 74: 132–138. Greenland, S. (Ed.) (1987). Evolution of Epidemiologic Ideas: Annotated Readings on Concepts and Methods. Epidemiology Resources Inc., Chapel Hill, NC.

High degree of heterogeneity Extremely relevant to man

Kahan, J.S. (1994). Medical Devices: Obtaining FDA Market Clearance. Parevel International, Waltham, MA. McMahon, B. and Pugh, T.F. (1970). Epidemiology, Principles and Methods. Little, Brown, Boston. MRC (1984). Expected Numbers in Cohort Studies. Medical Research Council Environmental Epidemiology Unit, Scientific Report No. 6, Southampton, U.K. Paddle, G.M. (1981). A strategy for the identification of carcinogens in a large, complex chemical company. In: R. Peto and M. Schneiderman (Eds.), Quantification of Occupational Cancer: Banbury Report 9. Cold Spring Harbor Laboratory, New York, pp. 177–186. Paddle, G.M. (1988). Epidemiology. In: D. Anderson and D.M. Conning (Eds.), Experimental Toxicology: The Basic Principles. Royal Society of Chemistry, London, pp. 436–456. Pell, S., O’Berg, M. and Karrh, B. (1978). Cancer epidemiologic surveillance in the Du Pont company. J. Occup. Med., 20: 725–740. Poole, C. (1986). Exposure opportunity in case-control studies. Am. J. Epidemiol., 123: 352–358. Rothman, K.J. (1986). Modern Epidemiology. Little, Brown, Boston. Schlesselman, J.J. (1982). Case-Control Studies: Design, Conduct, Analysis. Oxford University Press, New York. Schlesselman, J.J. and Stadel, B.V. (1987). Exposure opportunity in epidemiologic studies. Am. J. Epidemiol., 125: 174–178. Smith, A.H. (1988). Epidemiologic input to environmental risk assessment. Arch. Environ. Hlth., 43: 124–127. Spilker, B. (1995). Guide to Clinical Trials. Raven Press, New York. Utidjian, H.M.D. (1987). The interaction between epidemiology and animal studies in industrial toxicology. In: B. Ballantyne, (Ed.), Perspectives in Basic and Applied Toxicology. John Wright, Bristol, U.K., pp. 309–329. Waterhouse, J.A.H., Muir, C.J., Shanmugaratnam, K. and Powell, J. (Eds.) (1982). Cancer Incidence in Five Continents, Vol. IV. International Agency for Research on Cancer, Lyon (IARC Scientific Publication No. 42). Wegman, D.H. and Eisen, E.A. (1988). Epidemiology. In: B.S. Levy and D.H. Wegman (Eds.), Occupational Health. Recognizing and Preventing Work-related Disease. Little Brown, Boston, pp. 55–73.

18 Special Studies

The previous chapters in this book have primarily addressed tests which are applicable to all or broad ranges of medical devices. However, regulatory authorities (and responsible practice) have also established some specific testing requirements for a number of broadly utilized but somewhat specialized devices such as cardiovascular devices and prostheses (ISO, 1989a, 1989b, 1989c, 1989d, 1989e, 2000; AAMI, 1996), contact lenses and their solutions (CRDH, 1985, 1995a), and tampons (CDRH, 1995b). These evaluations require some specialized tests. In addition, there is one promulgated (but now infrequently used) test, the “mouse safety” or systemic injection test, which fits none of the previous classifications and needs to be briefly considered here. Many of these tests are designed to evaluate local tissue responses in the specialized tissues or uses that they have contact with.

Procedures used to evaluate cardiovascular devices in animals are essentially the same as those employed in the clinical setting. However, animal models permit continuous device monitoring and systematic controlled study of important variables. The test protocols recommended follow certain general guidelines. Thrombosis, thromboembolism, bleeding, and infection are the major deterrents to the use and further development of advanced cardiovascular prostheses. For devices with limited blood exposure (<24 hours), important measurements are related to the acute extent of variation of hematologic, hemodynamic and performance variables, gross thrombus formation, and possible embolism. With prolonged or repeated exposure (>24 hours), or permanent contact, emphasis is placed on serial measurement techniques that may yield information regarding the time course of thrombosis and thromboembolism, the consumption of circulating blood components, the development of intimal hyperplasia, and infection. In both of the above exposure and contact categories, assessment of the hemolysis is important. Thrombus formation may be greatly influenced by surgical technique, variable time-dependent thrombolytic and embolic phenomena, superimposed device infections, and possible alterations in exposed surfaces, for example, intimal hyperplasia and endothelialization. The consequences of the interaction of artificial surfaces with the blood may range from gross thrombosis and embolization to subtle effects such as accelerated consumption of hemostatic elements; the latter may be compensated or lead to depletion of platelets or plasma coagulation factors. Disturbances of organ function may occur due to blood/device interaction. For instance, kidney function and pulmonary function may be affected by activated blood coagulation and platelet/leukocyte/complement interactions. Platelet survival and plasma levels of the plateletspecific proteins PF-4 and β-TG may reflect the extent of platelet activation in vivo (and perhaps risk of thromboembolism) even in the absence of significantly elevated rates of platelet consumption. The template bleeding time is an index of in vivo platelet function; a prolonged value suggest thrombocytopenia or a qualitative platelet disorder, such as may occur during cardiopulmonary bypass. Measurements of FPA may indicate activation of intrinsic coagulation. Localization of thrombotic material by radionuclide imaging techniques using radio-labeled platelets has been demonstrated in studies of vascular grafts, valve prostheses, and other devices, both implants and externally communicating devices. In addition, duplex scanning and a careful examination of the explanted device can provide very useful information.

CARDIOVASCULAR DEVICES AND PROSTHESES Interactions which mainly affect cardiovascular devices and which may or may not have an undesirable effect on the subject are as follows: 1. Adsorption of plasma proteins, lipids, calcium, or other substances from the blood onto the surface of the device or absorption of such substances into the device. 2. Adhesion of platelets, leukocytes, or erythrocytes onto the surface of the device, or absorption of their components into the device. 3. Alterations in mechanical and other properties of the device. Interactions which have a potentially undesirable effect on the patient (animal or human) are as follows: 1. Activation of platelets, leukocytes, or other cells, or activation of the coagulation, fibrinolytic, complement, or other pathways, including immunotoxicity (immunosuppression, immunopotentiation, or immunomodulation). 2. Formation of thrombi on the device surface. 3. Embolization of thrombotic or other material from the device’s luminal surface to another site within the circulation. 4. Injury to circulating blood cells resulting in anemia, hemolysis, leukopenia, thrombocytopenia, or altered function of blood cells. 5. Injury to cells and tissues adjacent to the device. 6. Intimal hyperplasia or accumulation of other tissue on or adjacent to the device, resulting in reduced flow or affecting other functions of the device. 7. Adhesion and growth of bacteria or other infectious agents to or near the device. 243

244

Safety Evaluation of Medical Devices

The choice of an animal model may be restricted by size requirements, the availability of certain species, and cost. It is critical that the investigators be mindful of the physiological differences and similarities of the species chosen with those of the human, particularly those relating to coagulation, platelet functions and fibrinolysis, and the response to pharmacological agents such as anesthetics, anticoagulants, thrombolytic, and antiplatelet agents, and antibiotics. Because of species differences in reactivity and variable responses to different devices, data obtained from a single species should be interpreted with caution. Nonhuman primates such as baboons bear a close similarity in hematologic values, blood coagulation mechanism, and cardiovascular system to the human. An additional advantage of a nonhuman primate is that many of the immunologic probes for thrombosis developed for humans are suitable for use in other primates. These probes include PF-4, β-TG, FPA, TAT, and F1+2. The dog is a commonly used species and has provided useful information; however, device-related thrombosis in the dog tend to occur more readily than in the human, a difference which can be viewed as an advantage when evaluating this complication. The pig is generally regarded as a suitable animal model because of its hematologic and cardiovascular similarities to the human. The effect of the surgical implant procedure on results should be kept in mind and appropriate controls included.

mortality (Bisno and Waldnagel, 1994). A property common to all these biomaterials is the ease by which they are colonized by pathogenic and nonpathogenic microorganisms, often requiring immediate removal. Several methods have been devised to decrease the risk of foreign body associated infections. These include the use of meticulous hygienic precautions, the development of hydrophillic materials to minimize bacterial adhesion and impregnation with antiseptics, antibiotics and a host of other pharmaceutical products. Silver, in particular free silver ions, is well known for its powerful and broad-spectrum antimicrobial activity still allowing the independent use of therapeutic antibiotics. The investigation of the antimicrobial activity of implants containing silver as an antimicrobial agent is difficult because many silver compounds are poorly water soluble, resulting in low concentrations of silver ions released into the surrounding medium. Therefore, the antimicrobial efficacy of polymers impregnated with elementary silver cannot be tested by routine agar diffusion measurements. Like other procedures, the agar diffusion technique was also inappropriate for a simultaneous high-throughput screening prototype devices. Bechert et al. (2000) have proposed a very promising microplate based system for such evaluations.

Cannulae

The hemostatic response to cardiopulmonary bypass may be significant and acute. Many variables such as use of blood suction, composition of blood pump priming fluid, hypothermia, blood contact with air, and time of exposure influence test values. Emboli in outflow lines may be detected by the periodic placement of blood filters ex vivo, or the use of ultrasonic radiation or other noninvasive techniques. Thrombus accumulation can be directly assessed during bypass by monitoring performance factors such as pressure drop across the oxygenator and oxygen transfer rate. An acquired transient platelet dysfunction associated with selective alpha granule release has been observed in patients on cardiopulmonary bypass: the template bleeding time and other tests of platelet function and release are particularly useful. Complement activation is caused by both hemodialyzers and cardiopulmonary bypass apparatus. Clinically significant pulmonary leukostasis and lung injury with dysfunction may result. For these reasons, it is useful to quantify complement activation with these devices. Therapeutic apheresis equipment and devices for absorption of specific substances from the blood, because of their high surface-to-volume ratio, can potentially activate complement, coagulation, platelet, and leukocyte pathways. Examination of blood/device interactions should follow the same principles as for extracorporeal oxygenators and hemodialyzers.

Cannulae are typically inserted into one or more major blood vessels to provide repeated blood access. They are also used during cardiopulmonary bypass and other procedures. They may be tested acutely or chronically and are commonly studied as arteriovenous (AV) shunts. The use of cannulae appears to induce little alteration in the levels of circulating blood cells or clotting factors. Cannulae, like other indirect blood path devices, generally require less testing than devices in circulating blood.

Catheters and Guidewires Most of the tests considered under cannulae are relevant to the study of catheters and guidewires. The location or placement of catheters in the arterial or venous system can have a major effect on blood/device interactions. It is advised that simultaneous controls studies be performed using a contralateral artery or vein. Care should be taken not to strip off thrombus upon catheter withdrawal. In situ evaluation may permit assessment of the extent to which intimal or entrance site injuries contributed to the thrombotic process. Kinetic studies with radiolabeled blood constituents are recommended only with chronic catheters, but may be useful for imaging thrombus accumulation in vivo. Angiography and Doppler blood flow measurements may also be useful.

Extracorporeal Oxygenators, Hemodialyzers, Therapeutic Apheresis Equipment, and Devices for Absorption of Specific Substances from Blood

Anti-Infective Materials Implantable medical devices such as cannulae, catheters and stents are indispensable in the management of critically and chronically ill patients for the administration of electrolytes, drugs, parenteral nutrients, blood components or drainage of secretions and pus. Artificial heart valves, prosthetics, ceramics, metals and bone cements are now common implants or implant materials. All of these implants save human lives and enhance quality of life. At the same time they are the leading cause for millions of primary nosocomial bloodstream infections with substantial morbidity and

Ventricular-Assist Devices These devices may induce considerable alteration in various blood components. Factors contributing to such effects include the large foreign surface area to which blood is exposed, the high flow regimes, and the regions of disturbed flow such as turbulence or separated flow. Tests of such devices may include measurements of hemolysis; platelet and fibrinogen concentration; platelet survival; complement activation; and close monitoring of liver, renal, pulmonary, and central nervous system function. A detailed pathologic

Chapter 18:

examination at surgical retrieval is an important component of the valuation.

Heart Valve Prostheses Invasive, noninvasive, and in vitro hydrodynamic studies are important in the assessment of prosthetic valves. 2D and M mode echocardiography makes use of ultrasonic radiation to form images of the heart. Reflection from materials with difference acoustic impedances are received and processed to form an image. The structure of prosthetic valves can be examined. Mechanical prostheses emit strong echo signals and the movement of the occluder can usually be clearly imaged. However, the quality of the image may depend upon the particular valve being examined. Echocardiagraphy may also be useful in the assessment of function of tissue-derived valve prostheses. Vegetations, clots, and evidence of thickening of the valve leaflets are elucidated. Using conventional and color flow Doppler echocardiography, blood flow regurgitation can be identified and semi-quantified. Measurements of platelet survival and aggregation, blood tests of thrombosis and hemolysis, pressure and flow measurements, and autopsy of the valve and adjacent tissues are also recommended.

Vascular Grafts Both porous and nonporous materials can be implanted at various locations in the arterial or venous system. The choice of implantation site is determined largely by the intended use for the prosthesis. Patency of a given graft is enhanced by larger diameter and shorter length. A rule of thumb for grafts less than 4 mm ID is that the length should exceed the diameter by a factor of 10 (i.e., 40 mm for a 4-m m graft) for a valid model. Patency can be documented by palpation of distal pulses in some location and by periodic angiography. Ultrasonic radiation, MRI, and PET may also be useful. Results of serial radiolabeled platelet imaging studies correlated with the area of nonendothelialized graft surface in baboons. Radiolabeled platelets facilitate noninvasive imaging of mural thrombotic accumulations. Serial measurements of platelet count, platelet release constituents, fibrinogen/fibrin degradation products, and activated coagulation species are also recommended. Autopsy of the graft and adjacent vascular segments for morphometric studies of endothelial integrity and proliferative response can provide valuable information.

IVC Filters and Stents These devices can be studied by angiography and ultrasonic radiation. Other techniques useful for vascular graft evaluation are appropriate here as well.

TAMPONS Tampons occupy a unique niche in both commerce and medical device regulation. They are sold freely at any commercial outlet (supermarket, drug store, minimarket, etc.) without restriction or control, and they are used regularly by as many as thirty-eight million women in the United States (and many more overseas), but do require premarket review and approval as a device by the FDA. As a result of the toxic shock syndrome (TSS) scare of the early 1980s (see Chapter 19 for details on this) and continuing concerns as to their potential effects on vaginal microflora and on the induction

Special Studies

245

of TSS toxin. As a result, the Obstetrics–Gynecology Devices Branch of CDRH maintains separate testing requirements and guidelines for menstrual tampons (CDRH, 1995b). The unique aspects of these requirements from a safety point of view are that testing be performed to evaluate both the effects of new and revised product on vaginal microflora and to determine their potential to lead to the production of TSS toxin. These specific tests are presented later in this chapter. Also of concern is the potential for any potentially hazardous materials to be leached from the tampon into the vaginal environment. Though this is a general concern, there has recently been imposed a specific requirement to evaluate dioxins in the absorbent fibers of the tampon.

CONTACT LENSES AND THEIR SOLUTIONS Contact lenses also present a unique case for both medical device regulation and for safety assessment. Whether hard lenses or one of the varieties of disposable soft lenses, the device itself is provided by prescription from a licensed professional and is placed into intimate contact with the eyes of the user on a daily basis over a period of years. The solutions used to clean and disinfect the lenses, however, are brought as over-the-counter consumer products without control on their direct sale. Such disinfecting and cleaning solutions are considered and regulated as devices (as they are not supposed to have any direct therapeutic effect on the body and are essential for the use of the lenses themselves). Such solutions also have direct daily contact with the eyes of millions of individuals on a daily basis over a period of years. As such, they must be evaluated for both their potential local and systemic effects, and are regulated under specific guidelines from the Contact Lens Branch of the CDRH (1985, 1995a). The specialized test specifically mandated for their evaluation is the 21-day eye irritancy test in rabbits, as described later in this chapter. However, the component materials in such solutions must be carefully considered (and potentially evaluated experimentally) for systemic effects also (Hackett and Stern, 1991). Such systemic toxicity evaluations are not generally performed in the rabbit, but are conducted with ocular administration of test solutions.

SPECIFIC TESTS Embryotoxicity Device and labware (primarily disposable syringes and culture dishes) utilized in in vitro fertilization (IVF) clinics are required to be screened for embryotoxicity. Both the Guidelines for Human Embryology and Andrology Laboratories (1992, promulgated by the American Fertility Society) and the regulations of the Commission on Laboratory Accreditation of the College of American Pathologists require the performance of such tests to assess the media, materials, and supplies that will be used during human in vitro fertilization procedures. Three different test systems currently see some use, two forms of the mouse embryo test (one form that starts with a one-cell embryo and another that uses a two-cell embryo as a starting point) and the hamster sperm motility test. These in vitro embryo assays have been found to be effectively predictive of in vivo results (Neubert et al., 1986). The mouse embryo test has significant limitations as a predictive screen for assessing the quality of the culture

246

Safety Evaluation of Medical Devices

conditions used in human in vitro fertilization (Brigin et al., 1986). However, its use is required by the minimal standards for IVF programs presented by the American College of Obstetrics and Gynecology (ACOG), 600 Maryland Ave., S.W., Washington, D.C., 20024. There are numerous considerations for quality control and maintaining assay performance (Quinn and Whittingham, 1982; Rinehart et al., 1988; Roblero and Riffo, 1986; Saito et al., 1984; Wiley et al., 1986).

Mouse Embryo Assay Despite its limitations, mouse embryo culture is the only simple reproducible method available for testing the embryo toxicity of culture conditions used in human IVF. Until a more sensitive system is developed, therefore, this assay is an essential component of a quality control program in human IVF (Biggers et al., 1981; Gianaroli et al., 1986; Davidson et al., 1988). There are numerous basic technique manuals for this methodology (Gerrity, 1988), and what is described here is adapted from these. Materials. Chemicals and media include the following: 1. Pregnant mare serum gonadotropin (PMSG) (Diosynth Corporation, Chicago, IL; Gestyl) 2. Human chorionic gonadotropin (Sigma Chemical Company, St. Louis, MO; #CG-10) 3. Dulbecco’s phosphate buffered saline (Gibco Labs, Grand Island, NY; #450-1300) 4. Ham’s F-10 culture media (Gibco Labs, Grand Island, NY; #450-1200) 5. BWW media (optional) 6. Human serum albumin (Fraction V, Sigma Chemical Company, St. Louis, MO; #A-1653) Animals. F1 of C3B6 (Charles Riber, Wilmington, MA); male and female. This strain of mouse responds well to superovulation and will undergo embryo development from two-cell stage to blastocyst in culture in 72 hr. In selecting a strain of animals for use in this bioassay, it is essential to determine whether development from two-cell to blastocyst does occur in the strain selected (i.e., is there a two-cell block in the strain that is intended for use?). The presence of a two-cell block can be investigated by harvesting both two-cell and four- to eight-cell embryos from the strain of mouse in question by simply altering the embryo collection times. These embryos are then cultured under the usual conditions. If development of the four- to eight-cell embryos proceeds to the blastocyst stage while the two-cell embryos have stopped development, a two-cell block exists. If development is poor in both groups, a problem with the culture medium should be suspected. The response of the mouse strain to superovulation regimens is also important; a minimum of 30–40 embryos from each animal is the goal. Finally, expense and availability in each geographic area may be important (Pannaud et al., 1987). It may also be cost effective to use a less expensive strain of male mice. For a summary of these characteristics in various strains of mice, see Ackerman et al. (1983, 1984, 1985). Mice should be housed under a 12-hr light/dark cycle (lights on at 06:00) with standard laboratory chow and water available ad libitum. Female mice may be group-housed according to appropriate American Association for Accreditation of Laboratory Animal Care Standards (usually 4–5 a cage). Male mice must be individually housed. For best results in mating, always add female mice to the male cage, not the reverse. Use only sexually mature animals (6–8 weeks of age); avoid using

female animals that are aged (over 3 months). Replace all males every 4–5 months. Female mice may be randomly selected for injection without attention to the stage of the estrus cycle. All animals must acclimate to their new environment for one week after shipping. Gonadotropins (PMSG + hCG): 1. Prepare stock solutions at a concentration of 100 IU/mL in saline or distilled water. 2. Freeze in small aliquots convenient for a single use. Do not refreeze gonadotropin after thawing. 3. Store PMSG and hCG in a freezer at 70°C or in a freezer that is not self-defrosting or frost-free. The freeze/thaw cycle of a standard kitchen frost-free refrigerator will inactivate frozen gonadotropins rapidly. 4. Never filter sterilize gonadotropins; they bind tightly to filters leaving a biologically inactive (though sterile) solution. Animal Preparation: 1. To superovulate female animals, inject intraperitoneally with 7 IU of PMSG. For best results, inject animals between 3 and 6 p.m. 2. 48 hr after PMSG injection, inject animals with 7 IU of hCG. For best results, inject animals between 3 and 6 p.m. 3. Immediately after giving the hCG injection, place one female mouse into the cage of one male mouse. For best results, do not exceed two female mice per male mouse. 4. Time of sacrifice. Approximately 24 hr after hCG injection for 1-cell embryos. Approximately 36 hr after hCG injection for two-cell embryos. These times will vary somewhat with the strain of mice used. 5. For guidelines on how many female mice to use for each assay, see the section below entitled “Setup and Preparation.” It should be noted that frozen embryos are commercially available and may be used, though their baseline viability levels are less than those of fresh embryos. See Table 18.1 for a schedule summary. Set-up and Preparation: 1. Place 1 mL of prewarmed Dulbecco’s PBS in each of several 35-mm Petri dishes (one dish per mouse). Place on a slide warmer to maintain approximate 37°C before use. Prepare several extra dishes of PBS for filling the flush syringe. 2. Pre-equilibrate media and sera to be tested in incubator containing 5% CO2 in air, preferably overnight. 3. When testing media or sera, the following guidelines are useful. (These guidelines are illustrated in Table 18.2.) a. A control medium that has been previously tested must be used in addition to the test medium. Some laboratories make use of a standard medium that is Table 18.1

Schedule for Superstimulation of F1C3B6 Mice

PMSG injection (3–6 p.m.)

HCG injection (3–6 p.m.)

Recovery: 1-cell embryo (3–6 p.m.)

Recovery: 2-cell embryo (8–10 a.m.)

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

Wednesday Thursday Friday Saturday Sunday Monday Tuesday

Thursday Friday Saturday Sunday Monday Tuesday Wednesday

Friday Saturday Sunday Monday Tuesday Wednesday Thursday

Chapter 18:

Special Studies

247

Table 18.2 Example of Typical Setup for a Falcon 3047 Microtiter Plate to Test Five Different Treatments with Four Different Mice Sample 1-mL microtiter plate (Falcon 3047) for mouse embryo culture

Fat pad

Treatmenta Mouse numberb

1

2

3

4

5

1 2 3 4

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

Bursa Cut 2 Fimbria Cut 3

a

Treatments (sample to be tested) 1. Control medium—either BWW containing 3 mg/mL of human serum albumin (HAS) or Ham’s F-10 (previously tested Batch 100) + 3 mg/mL HAS. 2. Unknown Batch Ham’s F-10 (e.g., Batch 101) + 3 mg/mL HAS. 3. Unknown fetal cord serum (FCS) Batch 001 in Batch 100 Ham’s f-10. 4. Unknown FCS Batch 002 in Batch 100 Ham’s F-10. 5. Unknown FCS Batch 003 in Batch 100 Ham’s F-10, where Batch 100 = previously tested Ham’s F-10; Batch 101 = unknown new batch of Ham’s F-10. b All of the embryos from each mouse to be evenly split across the five treatments.

known to support mouse embryo development, such as BWW medium of Tyrode solution, for this purpose. The author recommends use of a previously tested batch of Ham’s F-10 medium that is less than 3 weeks old. b. When testing unknown (untested) sera, always test in a culture medium that has been previously tested. c. A protein source must be added to the control medium as well as to the medium to be tested (Caro and Trounson, 1984; Ogawa and Marrs, 1987). The protein source must be a previously tested batch of serum or human serum albumin (Fraction V); concentration: 3 mg/mL (this source of protein should not be included in medium which is used to screen untested sera). d. There is a high degree of mouse-to-mouse variability in the rate of fertilization of ova and in the rate and number of embryos to complete development. Therefore, all of the embryos from one mouse should be pooled and divided evenly among treatment conditions. e. Do not pool all embryos from all mice and then divide because of the large interanimal variation in development. If all the poor embryos are not evenly distributed across treatments, it may bias interpretation of the test results. Mice that yield greater than 25% fragmented, unfertilized, or one-cell embryos or embryos that have developed beyond the two-cell stage should be excluded from the testing procedure. f. For a sample of assay setup, see Table 18.2. For convenience, 1-mL microtiter plates (Falcon 3047) containing 1 mL of test media in each well may be used. These trays have four rows (A–D) and five columns (1–5). g. The setup of the test wells (the bioassay scheme) may be recorded. 4. Guidelines for deciding on number of mice to inject: a. Each mouse should yield 30–40 embryos. b. It is best to have at least five embryos per treatment; therefore, one mouse could test five treatments easily. Replicates are required so additional mice must be injected. c. Not all mice will superovulate. d. For the sample given in Table 18.2, five mice would have to be injected. At least five embryos from mouse 1 would be placed in wells A1 through A6. This process would be repeated with mouse 2 in row B and so on.

Ovary Bursa Oviduct

Cut 1 Uterine horn

Cervix

Figure 18.1 Schematic description of female mouse reproductive tract and surgical removal of oviducts.

e. When testing a new batch of plastic ware to determine if it is embryotoxic, always be certain to use embryos obtained from the same mouse grown in the same medium but placed in previously tested plastic ware as a control. f. Media and sera for IVF should be tested in the incubator used for IVF as a quality control measure to evaluate the entire embryo culture system (Arny et al., 1987; Barrach and Neubert, 1980). Removal of Oviducts: 1. Sacrifice animals by cervical dislocation. 2. Soak the animal with 70% alcohol to minimize hair contamination when the abdomen is opened. 3. Use coarse instruments to open the abdomen. Pneumothorax the animal. To accomplish this, make a midline cut from the pubic symphysis through the diaphragm using coarse scissors and forceps. Extend the cuts laterally to flap the skin back. 4. Using fine instruments, isolate the reproductive tract and identify the oviducts (refer to Figure 18.1). 5. When removing the oviduct, it is important to avoid handling or squeezing the oviduct itself. For ease of handling, first sever the uterine horn as shown in Figure 18.1, Cut 1; use this piece of tissue as a kind of handle for grasping the tissue. Next cut the oviduct between the ovary and the oviduct as shown in Figure 18.1, Cut 2. At this point, the oviduct is free from the mouse and is held by a piece of uterine horn. Hold this piece of tissue over a 30-mm Petri dish containing warm PBS. Holding the uterine horn, cut between the uterine horn and the oviduct as in Figure 18.1, Cut 3, allowing the oviduct to drop into the warm PBS.

248

Safety Evaluation of Medical Devices

6. Repeat with second oviduct. 7. For best results, flush these two oviducts before proceeding to sacrifice the next mouse. Do not pool oviducts from several mice in the same dish. Flushing the Oviducts: 1. Fill a 1-cc syringe with warm PBS and place a 30-gauge, 1-in. needle on the end. Sterile disposable needles of this size are available from Becton Dickinson and are convenient and inexpensive to use. The sharp end of this needle is a problem for some investigators, but practice can eliminate puncturing the sides of the oviduct. Some groups prefer to blunt the end of the needle using an emery board; these needles must be resterilized before use. Custom-made needles that have no bevel or point can also be ordered from several manufacturers but these needles are expensive and are not disposable. 2. Place the 35-mm Petri dish on the stage of the dissecting microscope and examine the oviduct grossly to identify the fimbria (refer to Figure 18.1) and the uterine end of the oviduct. 3. Carefully work the 30-gauge needle into the fimbriated end of the oviduct, using the forceps to grasp the oviduct. (Very fine watchmakers’ forceps will make handling simpler.) 4. Gently press the plunger on the syringe and watch the embryos flush out of the uterine end of the oviduct. If the needle is properly placed, the injection of medium should partially distend the oviduct. Flushing should require no more than 0. 1 mL of medium. The smaller the volume flushed, the easier will be the recovery of embryos once these have been flushed into the dish. 5. If the needle goes through the wall of the oviduct, withdraw it partially and attempt to pass the needle beyond the puncture site, grasping the oviduct distally; flush again. 6. It is also possible to flush from the uterine end of the oviduct. This end has a more muscular wall and some will prefer to handle the oviduct at this end. 7. Check the dish to ensure that the embryos have been flushed. 8. If none of the above techniques work, it is also possible to use the needle and forceps to break the oviducts allowing the embryos to spill out. This method is much more time consuming, however, and the number of embryos obtained may be reduced. 9. Repeat the procedure with the second oviduct. Pool the embryos from both oviducts. Separate out all embryos that are not two-cell embryos (including fragmented, one-cell embryos, and embryos at four-cell or beyond). When collecting two-cell embryos, if more than 25% of the embryos recovered are unfertilized, fragmented, one-cell or have developed beyond the two-cell stage, that mouse should not be used in testing and all of the embryos should be discarded. Place equal numbers of two-cell embryos in each treatment to be evaluated. 10. It is most convenient to handle the embryos using a Pasteur pipe pulled to a fine bore over a low flame and attached to a PiPump. 11. Place the microtiter plate containing the embryos in the incubator and repeat the procedures above with each mouse in sequence. 12. Record the total number of embryos in each well.

Monitoring and Analyzing Results: 1. Mouse embryos should be observed at 24-hr intervals after placing in culture (i.e., 24, 48, and 72 hr after collection). Embryos may be examined directly under the dissecting microscope in the test wells of the microtiter plate or in Petri dishes. The method used is a matter of personal preference. Petri dishes have a large surface area to scan in order to find and examine all of the embryos. While microtiter plates confine the embryos to a small area for observation, the optical properties of these dishes often make it difficult to see embryos which settle around the edges of the well. Flat-sided culture tubes have also been used for these purposes but necessitate a large volume of media and attention to adequate equilibration of the media. Finally, some investigators prefer to place the embryos on a slide with a coverslip; this method only permits observation of the embryos at the final observation time point. 2. The total number of embryos in each well should be verified and the state of development noted each day that the embryos are observed. Care must be taken to make these observations rapidly since the pH and temperature of the media will change while the embryos are outside of the incubator. Such changes may retard embryonic development. For research applications, it is particularly useful to see if the rate of development is altered and at what stage development stopped. For the strain of mice used here (F1C3B6), embryos should progress from two-cell to hatched blastocyst in 72 hr. Some laboratories prefer to make a single observation at 72 hr and record the percentage of total embryos that have progressed from two-cell to hatched blastocyst. If 75% of the embryos do not progress from two-cell to hatched blastocyst in 72 hr, the serum, media or other test substance should be considered less than optimal and, therefore, discarded. Since mouse embryos are quite hardy when grown in culture, conditions which regard development should not be taken lightly. Media that “comes close to passing” should be retested before use. This would be reflected at 72 hr after embryo collection. Inspection of results at 24 and 48 hr will indicate when development stopped. 3. Timing of development Stage of development

Time (h)

1 cells 2 cells 4–8 cells Morula Early blastocyst Hatched blastocyst

−12 0 (time placed in culture) 24 36 48 72

4. Returning to the example presented in Table 18.1, comparisons are within rows (control A1 is compared to treatments A2–A6). Comparison between rows are an indication of the interanimal variation. From a statistical viewpoint, this interanimal variation must be taken into account before valid statistical comparisons using pooled date between mice can be made. If examination of results from Mouse 1 indicate that in the control treatment (A1) only 50% of the embryos reached hatched blastocyst, this mouse should be dropped from the data set. (Please note that the purpose of this assay is to screen for embryo toxicity so a good control group is essential. Use of this assay for a research application would not permit dropping of data.)

Chapter 18:

If all controls A1, B1, C1, and D1 showed slowed development, however, the assay should be repeated. 5. The use of one-cell mouse embryos in addition to or even in place of two-cell embryos has been advocated (Quinn et al., 1984). These embryos are apparently more sensitive to adverse culture conditions than are two-cell embryos. There are also different requirements for embryo development from the one-cell to the two-cell stage than for two-cell and beyond, making their use a more rigorous quality control measure. Uses of the Bio-assay: All new batches of media should be tested using the mouse embryo bio-assay. Experience indicates that Ham’s F-10 medium will support mouse embryo development for up to a month after preparation, but no conclusions can be drawn concerning effects on human embryo development. Furthermore, the purity of the water (the largest single component of the medium) will decline with time. As a precaution, the medium should be discarded two weeks after preparation. Gloves worn during any surgery or manipulation of the embryos may also be a source of problems (Naz et al., 1986). All sera used in the IVF program should be tested in the mouse embryo bio-assay after heat inactivation and filter sterilization. Experience indicates that far more sera than media will be rejected due to bioassay results. Although results vary from center to center (Condon-Mahoney et al., 1985), experience has demonstrated that less than 5% of media tested will be discarded based on results of mouse embryo culture. However, careful screening of sera leads to rejection of 25–30% of sera. Programs that do not test sera might inadvertently expose human embryos to embryo toxic serum and compromise success rates. The mouse embryo bioassay is not an all-or-none test; a test sera may permit growth to a four-cell, eight-cell, or even unhatched four blastocyst stage. Obviously, the same sera used in the human IVF setting may yield seemingly normal embryonic development which actually halts after embryo transfer. The fact that human ova undergo fertilization and cleavage up to the point of transfer should not be taken as a sign that the sera are safe for embryos unless verified in the mouse serum. Testing serum drawn using the IVF stimulation cycle is often not possible because of time constraints. In addition, if clomid is used during ovulation induction, it has a direct embryo toxic effect (Laufer et al., 1983). If maternal serum is to be used as a protein source, it must be drawn in a cycle prior to the IVF cycle to permit adequate testing. All sera used for IVF that are not actual matched patient sera should be screened for HTLV-III, hepatitis surface antigen, and core antibody. In cases where IVF is used as a treatment of immunologic infertility or for infertility of unknown etiology, the patient’s own serum should not be used. Alternatives include the use of fetal cord serum or appropriately screened donor serum. The use of serum may not be necessary in human IVF. Caro and Trounson (1984) compared media without added protein or supplemented with albumin or fetal cord serum for the ability to sustain mouse embryo development in vitro. The resultant embryos, when transferred to pseudopregnant recipients, produced pregnancy rates significantly higher than in the group where no protein supplementation was used. The same authors recently did a similar study in a human IVF program (Caro and Trounson, 1986). In this report, 10% maternal serum in T6 culture medium was compared with T6 medium containing no protein. There was no

Special Studies

249

significant difference in pregnancy rate in the group where no protein was used to supplement the media. A similar result has been described by European investigators (Menezo et al., 1984; Feichtinger et al., 1986). Feichtinger’s group made use of commercially available Menezo B-1 medium. Thus far, these studies have not been done using Ham’s F-10, the culture medium most widely used in the United States. All new batches of plastic ware, tubing, etc., or any material that comes into contact with human eggs or embryos, should be tested for embryo toxicity. This is easily accomplished by placing known embryo-safe pretested medium in contact with the test surface (either by washing the medium or actually culturing embryos in it) and performing the assay as described. This is a useful method for testing the effectiveness of glassware washing techniques, cleansing of surgical instruments (aspiration needle), transfer catheters, and any other item that comes in contact with the human eggs, sperm, or embryos. Adverse effects of plastic ware on embryo development have also been described by Schiewe et al. (1984). In summary, when using this strain of mice, any condition that does not result in mouse embryo development from the two-cell to hatched blastocyst in 72 hr should be discarded or re-evaluated. The limitations of this assay are often overlooked. This assay can only detect conditions which are grossly and harshly embryonic. It cannot detect or differentiate growth promoting factors. Perhaps pregnancies in IVF programs are associated with the coincidental use of serum containing a growth promoting factor that cannot be discerned by the mouse embryo assay. The addition of protein to the medium may absorb toxic substances as may the amino acids present in Ham’s F-10 medium. Therefore, the medium may quench the system, dampening the sensitivity of the assay. It is possible, therefore, that mouse embryos will develop in conditions that might slow or stop human development. In the human RAF-ET procedure, embryos are transferred at an early stage of development. It is possible that these embryos stagger through one or two cleavage divisions after which they cease development. This same phenomenon often occurs in the mouse embryo system in the presence of embryotoxic substances, but only longer periods of observation allow detection of the problem. Many labs have had experience with mouse embryos growing in a variety of conditions that are less than optimal; these embryos appear to be relatively indestructible when grown in Ham’s F-10. Some laboratories formerly made use of sperm viability assays for testing culture media, sera, etc. However, human spermatozoa from normal men are even more resistant to adverse conditions than are mouse embryos. In addition, the end point of sperm viability is usually a subjective parameter (for example, motility) (Gorrill et al., 1991). This method, therefore, has been largely abandoned. Perhaps this method of media evaluation will be resurrected when computer-assisted, objective methods of measurement of sperm motility, velocity, and even lateral head displacement are more widely available. Finally, a hamster sperm viability assay has recently been devised which purports even greater sensitivity to embryotoxic factors and is simpler, more chemically defined, and more economical than the mouse bioassay. A bioassay system for embryo-derived platelet activating factor has been described for differentiating viable embryos. This assay may be useful for predicting success of embryo transfer (O’Neill et al., 1985). Frozen mouse embryos are now available commercially for use in IVF quality control.

250

Safety Evaluation of Medical Devices

Unfortunately, these embryos are frozen at the eight-cell stage and have therefore bypassed the more sensitive stages of development. Furthermore, to carry out adequate quality control testing weekly would be prohibitively expensive. In conclusion, there is a need for objective, sensitive, and reproducible methods for testing materials used in human IVF for both embryotoxic and growth promoting factors. The limitations of the mouse embryo bioassay are obvious; however, it is the standard of practice in the field and should be required in all settings. The hamster sperm motility assay should be considered as an alternative to the mouse embryo assay.

21-DAY EYE IRRITATION STUDY IN RABBITS This study is intended to identify the irritation potential of a test contact lens or lens disinfectant solution compared to that caused by a control lens or disinfectant solution. Sixteen young adult (2–3.5 kg) New Zealand white rabbits are randomized into two groups. Table 18.3 1.

Test and control lenses (or lenses soaked in test and control disinfectants) are inserted into the right eyes of each rabbit in the appropriate group (control or test) once daily for 21 consecutive days. The test or control contact lens will be carefully placed over the cornea of each rabbit. The lens will remain in place for 24 hr. After 24 hr, the contact lens will be removed, the eye examined, and a new lens placed on the cornea for 24 hr. This procedure will be repeated for 21 consecutive days. The treated eye of each animal is examined and scored for irritation of the cornea, iris, and conjunctiva pretest, and prior to each daily application of the contact lens and again 24 hr following the last application (Day 22). A hand held source of illumination will be used to aid in examining the eyes. Ocular responses will be graded according to the scale in Table 18.3. Additional signs will be described.

Anaylsis of Data Calculations. The primary eye irritation score for each rabbit is calculated from the weighted Draize scale. The method of calculation is indicated on the scale included herein.

Scale for Scoring Ocular Lesions (Draize et al., 1944)

Cornea A. Opacity: degree of density (area most dense taken for reading): No ulceration or opacity Scattered or diffuse areas of opacity (other than slight dulling of normal luster, details of iris clearly visible Easily discernible translucent area, details of iris slightly obscured Opalescent areas, no details of iris visible, size of pupil barely discernible Opaque cornea, iris not discernible through the opacity

0 1 2 3 4

B. Area of cornea involved: One-quarter (or less), but not zero Greater than one-quarter, but less than one-half Greater than one-half, but less than three-quarters Greater than three-quarters up to whole area Score = A × B × 5

1 2 3 4 Maximum total

2.

A. Values: Normal Folds above normal, congestion, swelling, circumcorneal injection (any or all of these or combination of any thereof), iris still reacting to light (sluggish reaction is positive) No reaction to light, hemorrhage, gross destruction (any or all of these)

80 0 1 2

Score = A × 5 Maximum total 3.

10

Conjunctivae A. Redness (refers to palpebral and bulbar conjunctivae excluding cornea and iris): Blood vessels normal Some blood vessels definitely hyperemic (injected) More diffuse, deeper crimson red, individual vessels not easily discernible Diffuse beefy red

0 1 2 3

B. Chemosis No swelling Any swelling above normal (includes nictating membranes) Obvious swelling with partial eversion of lids Swelling with lids about half closed Swelling with lids more than half closed

0 1 2 3 4

C. Discharge No discharge Any amount different from normal (does not include small amounts observed in inner canthus of normal animals) Discharge with moistening of the lids and hairs just adjacent to lids Discharge with moistening of the lids and hairs and considerable area around the eye Score = (A + B + C) × 2 Maximum total

0 1 2 3 20

Chapter 18:

Interpretation. The interpretation of the scores recorded is based on the number of rabbits with positive scores. A positive score is any score for opacity of units, or a score of two or more for redness or chemosis. Irritant. An irritant is a test article which under the conditions of this study causes a positive score in four to six rabbits with unwashed eyes (16 CFR 1500.3(c)(4)). Indeterminate. If positive scores are noted in two to three animals with unwashed eyes, irritation potential of the test article is not classified. Nonirritant. If positive scores are recorded in zero or one animal, the test article is classified as a nonirritant.

SYSTEMIC INJECTION TEST This test (USP, 2007) is designed to evaluate systemic responses to the extracts of materials under test following injection into mice. Long both an FDA requirement and mainstay for use in quality control of lots of devices, it is now neither because of careful consideration of the minimal benefit derived, despite the significant use of animals involved. The experimental design is outlined in Table 18.4.

Test Animal Use healthy, not previously used albino mice weighing between 17 and 23 g. For each test group use only mice of the same source. Allow water and food, commonly used for laboratory animals and of known composition, ad libitum.

Procedure Agitate each extract vigorously prior to withdrawal of injection doses to ensure even distribution of the extracted matter. However, visible particulates should not be injected intravenously. Inject each of the five mice in a test group with the test sample or the control blank as outlined in Table 18.4, except to dilute each gram of the extract of the test sample prepared with polyethylene glycol 400, and the corresponding blank with 4.1 volumes of sodium chloride injection to obtain a solution having a concentration of about 200 mg of polyethylene glycol per mL. Observe the animals immediately after injection, again 4 hr after injection, and then at least 24, 48, and 72 hr. If during the observation period none of the animals treated with the extract of the test sample shows a significantly greater biological reactivity than the animals treated with the control blank, the test sample meets the requirements of this test. If any animals treated with the test sample show only slight signs of biological reactivity, and not more than one

Table 18.4

Injection Procedure—Systemic Injection Test

Extract or blank Sodium chloride injection 1 in 20 solution of alcohol in sodium chloride injection Polyethylene glycol 400 Drug product vehicle (where applicable) Vegetable oil

Dose/kg

Routea Injection rate (µ/sec)

50 mL 50 mL

IV IV

100 100

10 g 50 mL

IP IV

— 100

50 mL 50 mL

IP IP

— —

a IV = intravenous (aqueous sample and blank); IP = intraperitoneal (oleaginous sample and blank).

Special Studies

251

animal shows gross symptoms of biological reactivity or dies, repeat the test using groups of 10 mice. On the repeat test, the material passes if all 10 animals treated with the test sample show no significant biological reactivity above that seen in the test blank animals.

PRODUCTION OF TOXIC SHOCK SYNDROME TOXIN-1 (TSST-1) Tampons that were found to be associated with the highest risk of toxic shock syndrome (TSS), including those made of polyester foam plus carboxymethylcellulose chips and polyacrylate rayon, differed from the other products principally in their chemical composition. A system has been developed for testing designs primarily to determine whether tampon fibers and other components have the propensity to increase production of TSST-1 (Parsonnet et al., 1996, Wong and Downs, 1989). Cultivation conditions and toxin analyses are such that small differences in TSST-1 can be detected, differences that may not be manifest under less hospitable conditions. All experiments are performed with the investigators blinded as to the nature of the test products in order to prevent bias in the interpretation of laboratory results; codes are not broken until results have been submitted to the manufacturer. All test products and controls are either cultivated simultaneously (in triplicate), or with one sample of each product (and control) represented in a series of consecutive experiments. Two types of controls and run with each experiment: culture medium alone, in order to determine whether the test products induce production of more TSST-1 than is made in the absence of a tampon, and incubation with tampons, that were associated with the highest risk of TSS and which have been shown to increase production of TSST-1 in a variety of test systems. This system is well suited to identify any adverse influence of tampon constituents on production of TSST-1 and accordingly, on risk of TSS.

Methods Materials Strain MN8, a TSST-1 producing strain of Staphylococcus aureus isolated from a patient with menstrual TSS, is used in these experiments. Bacteria are cultivated in medium consisting of 3% N-Z-amine A (Sheffield Products, Norwich, NY), a pancreatic digest of casein, plus 1% yeast extract (Difco Laboratories, Detroit), which is known to yield high levels of toxin (Wong and Downs, 1989). Tampons (with the investigators being blinded as to the chemical composition of constitutive fibers and additives) constitute the test material. Rely Regular tampons are used as controls. Each tampon is weighed and the mean weight of each type of tampon is calculated. As there was a narrow range of weights for each type of tampon, the mean weight of each type is used to calculate how much medium to use in cultivations. Tampons are sterilized by exposure to ethylene oxide using an Anprolene sterilizer.

Shake-Flask Cultivations Tampons are placed in sterile fashion into 250-mL Erlenmeyer flasks, to which are added appropriate volumes of culture medium so as to yield a constant weight/volume ratio of tampon to medium of 4%. All tampons are tested in triplicate in a single experiment except for Rely, which is run

252

Safety Evaluation of Medical Devices

in duplicate. Three flasks containing medium alone are inoculated as a control. Strain MN8 is suspended in medium to an optical density of 0.8 at 650 nm, which is intended to yield a bacterial concentration of approximately 1.0 × 109 cfu/mL at the start of incubation. Flasks are incubated in a shaking water bath under the following condition: 37.0°C, 90 revolutions per minute, and gas mixture of 5% carbon dioxide, 20% oxygen, and 75% nitrogen. This set of conditions has been shown to yield high levels of TSST-1 in vitro. After 20 hr, culture fluids are decanted from the flasks in order to recover all fluid not retained by the tampons. Experiments have shown that the decantate is also representative of fluid retained by the tampon, in terms of both bacterial numbers and concentration of toxin. Culture fluids are streaked onto blood agar plates to check for purity, and aliquots are removed for determination of bacterial numbers. Bacteria are then removed by centrifugation and the resultant supernatants are harvested for testing for TSST-1. TSST-1 is measured using a competitive ELSA, as described by Parsonnet et al. (1985).

Statistics Results from replicate samples are pooled for statistical analysis. Results of test tampons and Rely are compared with those of medium along by ANOVA, with statistical significance determined by Fisher’s protected least significant difference test (P, 0.05). Two analyses are run, one including all samples and one excluding Rely Regular tampons.

IN VITRO EVALUATION OF THE EFFECTS OF POTENTIAL TAMPON MATERIALS ON pH AND GROWTH OF VAGINAL MICROFLORA The role of the vaginal microflora, both as a protective element against and as a possible catalyst for the development of TSS, has received considerable attention during the last five years. The recognition that many cases of TSS are associated with menstruation, along with the widespread use of tampons in this country, has led to the proposal that this form of catamenial protection may alter the normal microflora during menstruation and thus potentiate the proliferation of Staphylococcus aureus and the production of TSST-1. Bartlett et al. (1977) evaluated both the quantitative and qualitative composition of the vaginal microflora in a study of a small group of women. These volunteers submitted self-obtained swabs during a single menstrual cycle, including the menstrual period. The most noteworthy finding was that the number of anaerobes remained relatively constant during the entire cycle, although there was considerable variation from an additional group of volunteers whenever, in the course of the menstrual cycle, they appeared for a vaginal examination. A quantitative assessment of the microflora for both groups combined revealed a 100-fold decrease in the number of aerobes in the last week of the menstrual cycle compared with the numbers obtained during menstrual flow. These findings were confirmed by the observations of Sautter and Brown (1980), who obtained vaginal swabs for a small number of volunteers several times during one month. They found that the variety of bacterial phenotypes isolated from any individual remained relatively constant, but that the numbers of each microbial species changed at different times during the cycle.

In a more extensive study conducted by Johnson et al. (1985), samples were obtained from 34 women by a vaginal wash method during both the menstrual and intermenstrual phase of the cycle. Quantitative assessment of the microflora during the menstrual cycle revealed that the total number of bacteria did not change significantly from one sample interval to another. Qualitative evaluation in the microflora did show significant variations during the menstrual cycle.

Method Six strains representative of bacterial species commonly isolated as part of the vaginal microflora of menstruating women are used to evaluate the effect of tampons on bacterial growth in vitro (Onderonk et al., 1977, 1987a, b). Three strains (Lactobacillus acidophilus, Prevotella bivia, and Peptostreptococcus anaerobius) are included as representatives of the obligate anaerobes, and three strains (Staphylococcus aureus, S. epidermidis, and E. coli) are included as representatives of the facultatie anaerobes present as part of the vaginal microflora. All six strains are grown in a defined medium simulating genital tract secretions under the anaerobic conditions associated with the human vagina. For each bacterial strain, two duplicate flasks containing approximately one-half of a tampon are inoculated and sampled as described previously. In addition, for each strain, two control flasks containing defined medium and no tampon material are inoculated with the same bacterial strain. Each flask was sampled at time 1 and after 24 and 48 hr of incubation and the aliquots of each sample are plated onto appropriate media for determination of total counts. All counts are recorded as the log10 cfu/mL of medium. Differences between the broth control and any of the test tampons is noted after 24 and 48 hr of incubation. Differences of less than one log in counts are within the normal variability for this growth system and are considered to be random observations rather than specific trends. One can also measure the effect on pH of the VDM following addition of test materials. Each measurement is performed in triplicate and the mean value plotted. The medium is adjusted to have a starting pH of either 5.0 or 6.0 to assess the buffering ability of the test substance on the test broth. One can also measure the effect that tampons containing various concentrations of new materials have on VDM broth pH when placed into the broth solutions. The pH is measured at one, two and three hours after tampon addition and the results expressed as the change from the starting pH of the broth prior to addition of the tampon. Each tampon is tested in triplicate. In order to assess the maximal potential effect of tampon addition, the VDM is adjusted to a starting pH of 5.0.

INFLAMMATORY RESPONSES TO BIOMATERIALS To test the inflammatory responses induced by biomaterials, the cage implant system (Jergenson et al., 1990), is most often used. This system allows the sequential examination of the exudate surrounding the implant without the need to sacrifice the animal. Utilizing this investigation too, Merchant et al. (1984) have shown that monocytes and macrophages follow the appearance of neutrophils in the exudate and that macrophages have a preferential adsorption to the biomaterial. Neutrophils have a relatively shorter presence (hours to days) and disappear from the exudate more rapidly than

Chapter 18:

macrophages, which remain for days to weeks or longer in the presence of a foreign body or infection. Eventually macrophages become the predominant cell type in the exudate, resulting in a chronic inflammatory response.

REFERENCES AAMI (1996). Biological Evaluation of Medical Devices, 5th Edn. AMI, Rockville, MD. Ackerman, S.B., Swanson, R.J., Adams, P.J. and Wortham, Jr., J.W.E. (1983). Comparison of strains and culture media used for mouse in vitro fertilization. Gamete Res., 7: 103–109. Ackerman, S.B., Swanson, F.J., Stokes, G.K. and Veeck, L.L. (1984). Culture of mouse preimplantation embryos as a quality control assay for human in vitro fertilization. Gamete Res., 9: 145–152. Ackerman, S.B., Stokes, G.L., Swanson, R.J., Tailor, S.P. and Fenwick, L. (1985). Toxicity testing for human in vitro fertilization programs. J. In Vitro Fert. Embryo Transfer, 2: 132–137. Arny, M., Nachtigall, L. and Quagliarello, J. (1987). The effect of preimplantation culture conditions on murine embryo implantation and fetal development. Fertil. Steril., 48: 861. Barrach, H.J. and Neubert, D. (1980). Significance of organ culture techniques for evaluation of prenatal toxicity. Arch, Toxicol., 45: 161–187. Bartlett, J.G., Onderdonk A.B., Drude, E., et al. (1977). Quantitative bacteriology of the vaginal flora. J. Infect. Dis., 136: 271–277. Bechert, T., Steinepicke, P. and Guggenbichler, J., (2000). A new method for screening anti-infective biomaterials. Nature Medicine, 6(8): 1053–1056. Biggers, J.D., Whitten, W.K. and Whittingham, D.G. (1981). The culture of mouse embryos in vitro. In: J.C. Daniel, Jr. (Ed.), Methods in Mammalian Embryology. W.H. Freeman and Company, San Francisco, pp. 86–116. Bisno, A.L. and Waldnagel, F.A. (1994). Infections Associated With Indwelling Medical Devices, 2nd Edn. American Society for Microbiology, Washington, DC. Brigin, J., Hogan, B. et al. (1986). Manipulating the Mouse Embryo. Cold Spring Harbor Laboratory. Caro, C.M. and Trounson, A. (1984). The effect of protein on preimplantation mouse embryo development in vitro. J. In Vitro Fert. Embryo Transfer, 1: 183–187. Caro, C.M. and Trounson, A. (1986). Successful fertilization, embryo development, and pregnancy in human in vitro fertilization (IVF) using a chemically defined culture medium containing no protein. J. In Vitro Fert. Embryo Transfer, 3: 215–217. CDRH (1985). Testing Guidelines for Class III Soft (Hydrophillic) Contact Lens Solutions. Contact Lens Branch, ODE, CDRH, DFA, July. CDRH (1995a). Premarket Notification (510(K)) Guidance Document for Contact Lens Care Products. Contact Lens Branch, ODE, CDRH, U.S. Food and Drug Administration, June 7. CDRH (1995b). Draft Guidance for the Content of Premarket Notification for Menstrual Tampons. Obstetrics-Gynecology Devices Branch, ODE, CDRH, U.S. Food and Drug Administration, May 25. Condon-Mahoney, M., Wortham, J.W.E., Jr., Bundren, J.C., Witmyer, J. and Shirley, B. (1985). Evaluation of human fetal cord sera, Ham’s F-10 medium, and in vitro culture materials with a mouse in vivo fertilization system. Fertil. Steril., 44: 521–525. Davidson, A., Vermesh, M., Lobo, R.A. and Paulson, R.J. (1988). Mouse embryo culture as quality control for human in vitro fertilization: the one-cell versus the two-cell model. Fertil. Steril., 49: 516. Draize, J.H., Woodard, G. and Calvery, H.O. (1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membrane. J. Pharm. Exp. Ther., 82: 377–390. Feichtinger, W., Kemeter, P. and Menezo, Y. (1986). The use of synthetic culture medium and patient serum for human in vitro fertilization and embryo replacement. J. In Vitro Fert. Embryo Transfer, 3: 87–92. Gerrity, M. (1988). Mouse embryo culture and bioassay. In: E.P. Wolf, (Ed.), In Vitro Fertilization and Embryo Transfer: A Manual of Basic Techniques. Plenum Press, New York, pp. 57–76.

Special Studies

253

Gianaroli, L. Seracchioli, R., Ferraretti, A.P., Trounson, A., Flamigni, C. and Bovicelli, L. (1986). The successful use of human amniotic fluid for mouse embryo culture and human in vitro fertilization, embryo culture and transfer. Fertil. Steril., 46: 907. Gorrill, R., Rinehart, J., Tamhane, A.C. and Gerrity, M. (1991). Comparison of the hamster sperm motility assay to the mouse one-cell and two-cell embryo bioassays as quality control tests for in vitro fertilization. Fert. Ster., 55(2): 345–354. Guidelines for Human Embryology and Andrology Laboratories (1992). Fert. Ster. Supplement 1, 58: 4, pp. 1A–16S. Hackett, D.W. and Stern, M.E. (1991). Preclinical toxicology/safety considerations in the development of opthalmic drugs and devices. In: D.W. Hobson (Ed.), Dermal and Ocular Toxicology. CRC Press, Boca Raton, FL. ISO (1989a). Cardiovascular Implants—Cardiac Valve Prostheses. ISO 5840-1. ISO (1989b). Cardiac Pacemakers—Part 1: Implantable Pacemakers. ISO 5840-1. ISO (1989c). Cardiovascular Implants—Tubular Vascular Prostheses— Part 1: Synthetic Vascular Prostheses. ISO 7198-1. ISO (1989d). Cardiovascular Implants—Tubular Vascular Prostheses— Part 2: Sterile Vascular Prostheses of Biological Origin—Specification and Methods of Test. ISO 7198-2. ISO (1989e). Cardiovascular Implants and Artificial Organs—Blood–Gas Exchangers (Oxygenators). ISO 7199. ISO (2000). Cardiac Pacemakers—Part 3: Low-profile Connectors (IS-1) for Implantable Pacemakers. ISO 5841-3. Jergensen, C.H., Huber, C.H., Zimmerman, T.P. and Worlber, G. (1990). 3-Deazaadenosine inhibits leukocyte adhesion and ICAM-1 biosynthesis in tumor necrosis factor-stimulated human endothelial cells. J. Immunol., 144: 653. Johnson, S.R., Petxold, C.R. and Galask, R.P. (1985). Qualitative and quantitative changes of the vaginal microbial flora during the menstrual cycle. Am. J. Repro. Immunol. Microbiol., 9: 1–5. Laufer, N., Pratt, B.M., DeCherney, A.H., Naftolin, F., Merino, M. and Merkert, C.L. (1983). The in vivo and in vitro effects of clomiphene citrate on ovulation, fertilization, and development of cultured mouse oocytes. Am. J. Obstet. Gynecol., 147: 633–639. Menezo, Y., Testart, J. and Perrone, D. (1984). Serum is not necessary in human in vitro fertilization, early embryo culture, and transfer. Fertil. Steril., 42: 750–755. Merchant, R., Anderson, J.M., Phua, K. and Hiltner, A. (1984). In vivo biocompatibility studies. II. Biomer: preliminary cell adhesion and surface characterization studies. J. Biomed. Mater. Res., 18: 309. Naz, R.K., Janousek, J.T., Moody, T. and Stillman, R.J. (1986). Factors influencing murine embryo bioassay: effects of proteins, aging of medium, and surgical glove coatings, Fertil. Steril., 46: 914. Neubert, D., Blankenburg, G., Chahoud, I. et al. (1986). Results of in vivo and in vitro studies for assessing prenatal toxicity. Environ. Hlth. Persp., 70: 89–103. Ogawa, T. and Marrs, R.P. (1987). The effect of protein supplementation on single-cell mouse embryos in vitro. Fertil. Steril., 47: 156. Onderdonk, A.B., Zamarchi, G.R., Rodriguez, M.L., Hirsch, M.L., Munoz, A. and Kass, E.H. (1987a). Quantitative assessment of vaginal microflora during the use of tampons of various compositions. Appl. Environ. Microbiol., 53: 2774–2778. Onderdonk, A.B., Zemarchi, G.R., Rodriguez, M.L., Hirsch, M.L., Munoz, A. and Kass, E.H. (1987b). Qualitative assessment of vaginal microflora during use of tampons of various compositions. Appl. Environ. Microbiol., 53: 2779–2784. Onderdonk, A.B., Polk, B.F., Moon, N.E., Goren, B. and Bartlett, J.G. (1977). Methods for quantitative vaginal flora studies, Am. J. Obstet. Gynecol., 122: 777–781. O’Neill, C., Pike, I.L., Porter, R.N., Gidley-Baird, A.A., Sinosich, M.H. and Saunders, D.N. (1985). Maternal recognition of pregnancy prior to implantation: methods of monitoring embryonic viability in vitro and in vivo. Ann. N. Y. Acad. Sci., 442: 429–439. Pannaud, J., Reme, J.M., Monrozies, X., Favrin, S., Sarramon, M.F. and Pontonnier, G. (1987). Mouse system quality control is necessary before the use of new material for in vitro fertilization and embryo transfer. J. In Vitro Fert. Embryo Transfer, 4: 56.

254

Safety Evaluation of Medical Devices

Parsonnet, J., Mills, J.T., Gillis, Z.A. and Pier, B.G. (1985). Competitive, enzyme-linked immunosorbent assay for toxic shock syndrome toxin-1. J. Clin. Microbiol., 22: 26–31. Parsonnet, J., Modern, P.A. and Giacobbe, K.D. (1996). Effect of tampon composition on the production of toxic shock syndrome toxin-1 by Staphylococcus aureus in vitro. J. Infect Dis., 173. Quinn, P. and Whittingham, D. G. (1982), Effect of fatty acids on fertilization and development of mouse embryos in vitro. J. Androl., 3: 440. Quinn, P., Warners, G.N., Kerin, J.F. and Kirby, C. (1984). Culture factors in relation to the success of human in vitro fertilization and embryo transfer. Fertil. Steril., 41: 202–209. Rinehart, J.S., Bavister, B.D. and Gerrity, M. (1988). Quality control in the in vitro fertilization laboratory: comparison of bioassay systems for water quality. J. In Vitro Fert. Embryo Transfer, 5: 335. Roblero, L.S. and Riffo, M.D. (1986). High potassium concentration improves preimplantation development of mouse embryos in vitro. Fertil. Steril., 45: 412.

Saito, H., Berger, T., Mishell, D.R. and Marrs, R. P. (1984). Effect of variable concentration of serum on mouse embryo development. Fertil. Steril., 41: 460. Sautter, R.L. and Brown, W.J. (1980). Sequential vaginal cultures from normal young woman. J. Clin. Microbiol., 11: 479–484. Schiewe, M.C., Schmidt, P.M., Bush, M. and Wildt, D.E. (1984). Effect of absorbed/retained ethylene oxide in plastic culture dishes on embryo development in vitro. Theriogenology, 21: 2160 (abstract). USP (2007). Systemic injection test. In: United States Pharmacopoeia XXX. United States Pharmacopoeia Convention, Rockville, MD. Wiley, L.M., Yamami, S. and Van Muyden, D. (1986). Effect of potassium concentration, type of protein supplement, and embryo density on mouse preimplantation development in vitro. Fertil. Steril., 45: 111. Wong, A.C.L. and Downs, S.A. (1989). Investigation by improved syringe method of effect of tampons on production in vitro of toxic shock syndrome toxin-1 by Staphylococcus aureus. J. Clin. Microbiol., 27: 2482–2487.

19 Toxicokinetics in Biomaterial and Device Safety Evaluation

INTRODUCTION

(v) Relationship between the age of the animal and the kinetics of the test compound. (vi) Relationship between the dose regimens of the test compound used in the toxicity studies and those employed clinically in man.

Among the cardinal principle of toxicology is that the means by which an agent comes in contact with or enters the body (i.e., the route of exposure or administration) does much to determine the nature and magnitude of its effects. Accordingly, an understanding of the nature of device/tissue contact and their implications for absorption is essential. The fundamental safety factor is the ratio between what levels in the plasma (or at the target organ tissue) cause adverse effects to those levels which are present in the use of the device. Biocompatability studies usually involve a control group of animals (untreated and/or formulation treated) and usually two or more treated groups with different levels of exposure to the device of interest with exposure approximately that used in man (as closely as possible). Occasionally there also may be recovery groups to determine if any observed effects are reversible (and if so, to what extent) In most instances the high dose level is expected to elicit some toxic effects in the animals, often expressed as decreased food consumption and/or below-normal bodyweight gain, and has been selected after consideration of earlier data, perhaps from dose range-finding studies, or at least to dose as high as possible by the intended route. The other two dose levels are anticipated not to cause toxic effects. Generally, but not always (e.g., non-steroidal anti-inflammatory drugs in rodents), the “low” dose level is a several-fold multiple of the expected human therapeutic or exposure level (preferably five-fold for non-rodent and ten-fold for rodents). However, without knowing the true relationship of these dose levels to each other with respect to the absorption, distribution and elimination of the new chemical entity as reflected by its pharmacokinetics, it is difficult to see how meaningful extrapolations concerning safety margins can be made from the toxicity data obtained. Also, without pharmacokinetic data from the positive control group, its inclusion is of limited value and the results obtained could lead to erroneous conclusions. Toxicokinetic studies can provide information on several aspects, knowledge of which greatly facilitates assessment of safety of the chemical entity. Six such aspects can be mentioned.

ISO guidelines (AAMI, 2003) dictate a clearly defined set of objectives for toxicokinetic studies: ●

●

Primary: To describe the systemic exposure achieved in animals and its relationship to dose level and the time course of toxicity studies. Secondary: (1) To relate the exposure achieved in toxicity studies to toxicological findings and contribute to the assessment of the relevance of these findings to clinical safety. (2) To support the choice of species and treatment regimen in nonclinical toxicity studies. (3) To provide information which, in conjunction with the toxicity findings, contributes to the design of subsequent nonclinical toxicity studies.

These data may be obtained from all animals on a toxicity study, or from representative subgroups, or from satellite groups, or from separate studies. If toxicology can be described as being the study of the effects of a chemical on an organism, metabolism can be described as the opposite—the effects of the organism on the chemical. Metabolism refers to a process by which a drug (xenobiotic) is chemically modified by an organism. It is part of the overall process of disposition of xenobiotic (ADME)— the process by which a chemical gains access to the inner working of an organism (absorption), how it moves around inside an organism (distribution), how it is changed by the organism (metabolism) and how it is eventually eliminated from the organism (elimination). The EPA definition of biotransformation or metabolism is “… the sum of processes by which a xenobiotic (foreign chemical) is handled by a living organism.” The mathematical formulae used to describe and quantify these processes are collectively known as pharmacokinetics. The EPA definition of pharmacokinetics is “… quantitation and determination of the time course and dose dependency of the absorption, distribution, biotransformation and excretion of chemicals.” The acronym ADME has been used to describe the multifaceted biological process. The term metabolism has also come into common jargon to describe the entire process. This science has long played a central role in pharmaceutical development but has played less a role in the development of other types of products. The purpose of this chapter is to introduce the basic concepts of ADME and practices of studies conducted to study it, as

(i) Relationship between the exposure levels used and the relative extent of absorption of the test compound. (ii) Relationship between the protein-binding of the test compound and the dose levels used. (iii) Relationship between pharmacological or toxicological effects and the kinetics of any leachates compound. (iv) Effect of repeated doses on the kinetics of the test compound.

255

256

Safety Evaluation of Medical Devices

described the regulations under EPA and OECD which require such data for non-pharmaceutical products and to give some real world examples.

relates to toxicology. First, however, one should consider the fundamental terminology used in pharmacokinetic studies (Table 19.1).

REGULATIONS

Absorption

FDA believes that data from studies on the adsorption, distribution, metabolism, and excretion of a chemical can provide insight into mechanisms of toxicity and are essential in the design and evaluation of results from other toxicity studies. Such data should be provided for all drugs and significant impurities. Recommendations for obtaining data on the metabolism and toxicokinetics of these substances are presented in AAMI guidelines (2003). In general, it is required that this information be obtained as part of initial and subsequent repeat dose studies with a drug.

PRINCIPLES An understanding of the design and analysis of toxicokinetic studies requires a broad understanding of the underlying concepts and principles inherent in the ADME process and in our current technology for studying such. Each of these four principle areas is overviewed from a practical basis as it Table 19.1

Fundamental Terms Used in Pharmacokinetic Studies

Absolute bioavailability Absorption Accumulation

Analyte Area under curve (AUC) Bioavailability

Biotransformation

Clearance Cmax Concomitant toxicokinetics Disposition Distribution Dose proportionality Enterohepatic circulation Enzyme induction Enzyme inhibition Excretion Exposure

Absorption describes the process by which a chemical crosses a biological membrane to gain access to the inner workings of an organism. For mammals, this process results in the entry of the chemical into the blood stream, or systemic circulation. In this case the process is also called systemic absorption. Pharmaceutical products, procedures and devices, such as hypodermic needles or catheters, can be used to by-pass biological barriers. Other products gain access to the systemic circulation via the oral, dermal, buccal or inhalatory route of administration. For a material to be toxic (local tissue effects are largely not true toxicities by this definition), the first requirement is that it be absorbed into the organism (for which purpose being in the cavity of the gastrointestinal (GI) tract does not qualify). Most pharmaceuticals are intended to gain such access. There are characteristics which influence absorption by the different routes, and these need to be understood by any person trying to evaluate and/or predict the toxicities of different moieties. Some key characteristics and considerations are summarized below by route.

The bioavailability of a dosage form relative to an intravenous administration. The process by which a xenobiotic and its metabolites are transferred from the site of absorption to the blood circulation. The progressive increase of chemical and/or metabolites in the body. Accumulation is influenced by the dosing interval and half-life of the chemical. The process can be characterized by an “accumulation factor,” which is the ratio of the plasma concentration at steady state to that following the first dose in a multiple dosing regimen. The chemical entity assayed in biological samples. The concentration of chemical and/or metabolites in the blood (or plasma/serum) integrated over time. This is typically considered the best indicator of exposure. The rate and extent to which a xenobiotic entity enters the systemic circulation intact, following oral or dermal administration. It is sometimes expanded to include therapeutically active metabolites. Also known as the comparative bioavailability. The process by which a xenobiotic is structurally and/or chemically changed in the body by either enzymatic or non-enzymatic reactions. The product of the reaction is a different composition of matter or different configuration than the original compound. The volume of biological fluid which is totally cleared of xenobiotic in a unit time. The maximum mean concentration of the chemical in the plasma. Also known as the peak plasma concentration. Toxicokinetic measurements performed in the biocompatability study, either in all animals or in representative subgroups or in satellite groups. All processes and factors which are involved from the time a chemical enters the body to the time when it is eliminated from the body, either intact or in metabolite form. The process by which an absorbed xenobiotic and/or its metabolites partition between blood and various tissues/organs in the body. The relationship between doses of a chemical and measured parameters, usually including tests for linearity. The process by which xenobiotics are emptied via the bile into the small intestine and then reabsorbed into the hepatic circulation. The increase in enzyme content (activity and/or amount) due to xenobiotic challenge, which may result in more rapid metabolism of a chemical. The decrease in enzymatic activity due to the effect of xenobiotic challenge. The process by which the administered compound and/or its biotransformation product(s) are eliminated from the body. Exposure is represented by pharmacokinetic parameters demonstrating the local and systemic burden on the test species with the test compound and/or its metabolites. The area under the matrix level concentrationtime curve (AUC) and/or the measurements of matrix concentrations at the expected peak-concentration time Cmax, or at some other selected time C(time), are the most commonly used parameters. Other parameters might be more appropriate in particular cases.

Chapter 19:

First-order kinetics First-pass effect Flux Half-life Hepatic clearance Kel

Lag time Metabolite characterization Metabolite identification Metabolite profile Monitor Nonlinear kinetics (saturation kinetics) Presystemic elimination Profile Protein binding Relative bioavailability Renal clearance Satellite Steady state Support

Tmax Total clearance Toxicokinetics Validate

Volume of distribution (Vd)

Toxicokinetics in Biomaterial and Device Safety Evaluation

257

Kinetic processes, the rate of which is proportional to the concentration. The phenomenon whereby xenobiotics may be extracted or metabolized following enteral absorption before reaching the systemic circulation. Term (that takes area into consideration) used to describe the movement of a chemical across a barrier. Most typically used to describe the absorption of a chemical across the skin as ug/cm2/hr. The time elapsed for a given chemical entity concentration or amount to be reduced by a factor of two. The rate of total body clearance accounted for by the liver. The elimination constant for a chemical in plasma. Typically calculated using the formula Kel = −ln[10] × b where b is the slope of the linear regression line of the log of the mean plasma concentrations vs. time from the tmax to 24 hours. The interval between compound administration and when the compound concentration is measurable in blood. The determination of physiochemical characteristics of the biotransformation product(s). The structural elucidation of the biotransformation product(s). The chromatographic pattern and/or aqueous/nonaqueous partitioning of the biotransformation products of the administered compound. To take a small number of matrix samples (e.g., 1 to 3) during a dosing interval to estimate C(time) and/or Cmax. Kinetic processes, the rate of which is not directly proportional to the concentration. The loss of that portion of the dose that is not bioavailable. This would include, among others, loss through intestinal and gut-wall metabolism, lack of absorption, and first-pass hepatic metabolism. To take (e.g., 4–8) matrix samples during a dosing interval to make and estimate of Cmax and/or C(time) and area under matrix concentration time curve (AUC). The complexation of a xenobiotic and/or its metabolite(s) with plasma or tissue proteins. The bioavailability relative to a reference or standard formulation or agent. The rate of total body clearance accounted for by the kidney. Its magnitude is determined by the net effects of glomerular filtration, tubular secretion and reabsorption, renal blood flow, and protein binding. Groups of animals included in the design and conduct of a toxicity study, treated and housed under conditions identical to those of the main study animals, but used primarily for toxicokinetics. An equilibrium state where the rate of chemical input is equal to the rate of elimination during a given dose interval. In the context of a toxicity study—to ratify or confirm the design of a toxicity study with respect to pharmacokinetic and metabolic principles. This process may include two separate steps: (a) Confirmation using toxicokinetic principles that the animals on a study were exposed to appropriate systemic levels of the administered compound and/or its metabolite(s). (b) Confirmation that the metabolic profile in the species used was acceptable, data to support this will normally be derived from metabolism studies in animals and in humans. The sampling time point at which Cmax occurs. The volume of biological fluid totally cleared of xenobiotic per unit time and usually includes hepatic clearance and renal clearance. The study of the kinetics of absorption, distribution, metabolism and excretion of toxic or potentially toxic chemicals. In the context of an analytical method—to establish the accuracy, precision, reproducibility, response function and the specificity of the analytical method with reference to the biological matrix to be examined and the analyte to be quantified. A hypothetical volume of body fluid into which the chemical distributes. It is not a “real” volume, but is a proportionality constant relating the amount of chemical in the body to the measured concentration in blood or plasma.

A. Oral and rectal routes (gastrointestinal tract) 1. Lipid-soluble compounds (nonionized) are more readily absorbed than water-soluble compounds (ionized). a. Weak organic bases are in the nonionized, lipidsoluble form in the intestine and tend to be absorbed there. b. Weak organic acids are in the nonionized, lipidsoluble form in the stomach and one would suspect they would be absorbed there, but the intestine is more important because of time and area of exposure. 2. Specialized transport systems exist for some moieties: sugars, amino acids, pyrimidines, calcium, and sodium. 3. Almost everything is absorbed—at least to a small extent (if it has a molecular weight below 10,000). 4. Digestive fluids may modify the structure of a chemical. 5. Dilution increases toxicity because of more rapid absorption from the intestine, unless stomach contents bind the moiety.

6. Physical properties are important—for example, dissolution of metallic mercury is essential to allow absorption. 7. Age—neonates have a poor intestinal barrier. 8. Effect of fasting on absorption depends on the properties of the chemical of interest. B. Inhalation (lungs) 1. Aerosol deposition a. Nasopharyngeal—5 µm or larger in man, less in common laboratory animals. b. Tracheobronchiolar—1–5 µm c. Alveolar—1 µm 2. If a solid, mucociliary transport may serve to clear from lungs to GI tract. 3. Lungs are anatomically good for absorption. a. Large surface area (50–100 m2). b. Blood flow is high. c. Close to blood (10 µm between gas media and blood).

258

Safety Evaluation of Medical Devices

4. Absorption of gases is dependent on solubility of the gas in blood. a. Chloroform, for example, has high solubility and is all absorbed; respiration rate is the limiting factor. b. Ethylene has low solubility and only a small percentage is absorbed—blood flow limited absorption. c. Parenteral routes. d. Dermal routes. As a generalization, there is a pattern of relative absorption rates which extends between the different routes that are commonly employed. This order of absorption (by rate from fastest to slowest and, in a less rigorous manner, in degree to absorption from most to least) is iv > inhalation > im > ip > sc > oral > id > other dermal. Absorption (total amount and rate), distribution, metabolism, and species similarity in response are the reasons for selecting particular routes in toxicology. In acute studies, however, these things are rarely known to us. So the cardinal rule for selecting routes of use in acute testing is to use those routes which mirror the intended route for human exposure. If this route of human exposure is uncertain, or if there is the potential for either a number of routes or the human absorption rate and pattern being greater, then the common practice becomes that of the most conservative approach. This approach stresses maximizing potential absorption in the animal species (within the limits of practicality) and selecting from among those routes commonly used in the laboratory that which gets the most material into the animal’s system as quickly and completely as possible to evaluate the potential toxicity. In general, chemicals cross biological barriers by one of three mechanisms: active transport, facilitative transport, and passive transport. In active transport, the chemical is specifically recognized by the organism, which then expends energy to take the chemical up, even against a concentration gradient. In facilitative transport, the organism produces a carrier molecule which reacts with the target molecule to form a complex which more easily traverses the membrane, but no energy is expended to take up the complex. Such complexes do not flow

against a concentration barrier. The simplest mechanism is passive transfer or diffusion. Here, a chemical flows down a concentration gradient (from high concentration to a lower concentration) and must passively (no energy expended by organism) cross a biological membrane. Passive transfer or diffusion is the most common (if not the only) mechanism involved in the absorption of the vast majority of chemicals in commerce. The other mechanisms involved in absorption will not be further discussed here. Drugs in solution have a natural tendency (more rigorously defined by the laws of thermodynamics) to move down a concentration gradient. That is to say, the individual molecules of solute tend to move from a region of high concentration toward regions of lower concentration. Also, the movement of a chemical across a permeable barrier, such as a biological membrane, is a process called diffusion, as illustrated by Figure 19.1. For most products, these biological barriers are either the wall of the gastrointestinal tract, the lining of the pulmonary system, and/or the skin. Absorption from the GI tract is controlled by a variety of factors. These include the acid/base characteristics of the chemical (described as the pKa), the solubility, the nature of the delivery (e.g., diet vs. gavage), the nature of any vehicle (suspensions vs. solution, or aqueous vs. nonaqueous), and the gastrointestinal tract of the species under study. Ionized or charged organic moieties do not readily pass through the lipophilic cell membranes of the epithelial cells that line the GI tract. Thus more acidic molecules tend to be more readily absorbed from the stomach while more alkaline materials tend to be absorbed from the small intestine. This is because at the acidic pH of the stomach, acidic chemicals tend to be nonionized. More alkaline chemicals tend to be more ionized in the stomach and less ionized in the gut. The equilibrium reaction for acidic dissociation can be represented by the equation: X–C–OH + H2O O

X–C–O + H3O+ O

Oral administration Sublingual administration Buccal cavity

Venous return from buccal cavity Hepatic vein

Stomach Bile duct Liver Intestine

Portal vein

Rectum

Lymphatics

Vena cava

Venous return from rectum Rectal administration

Figure 19.1

Passage of chemical moieties from the gastrointestinal tract into the bloodstream, shown in a diagrammatic fashion.

Chapter 19:

Like all chemical equations, this one has an equilibrium constant. The discussion of basic chemistry is outside the purview of this book. Readers who may need a refresher are referred to Tse and Jaffe (1991). For every chemical, a pKa can be calculated, based on equilibrium constant, which represents the proportion of ionized and unionized material in solution. The lower the pKa of a chemical the more likely it is to be nonionized in the stomach.

Absorption from the Pulmonary System Of the three routes discussed here, absorption from the pulmonary system is perhaps the most rapid. Systemic absorption of inhaled materials is highly dependent on the physical properties of the inhaled materials which dictate how easily the materials reach the alveoli of the deep lung. Gases and vapors easily penetrate into the deep lung. For mists and dusts, absorption will be highly dependent on particle size. In general, the larger the particles, the less they will penetrate the pulmonary system. The term impaction describes the deposition of particles in the respiratory tract. Particles of less than 0.2 µm are preferentially deposited in the pulmonary portion of the respiratory system and particles over 2 µm do not reach the alveolar epithelium in great number. Particles from 1 to 4 µm tend to be distributed over the length of the system and particles over 4 µm tend to be deposited in the nasal region. Aerosolized particles of greater than 20 µm do not commonly occur in nature. Tidal volume will also influence impaction. In general, the larger the tidal volume, and thus the more forceful the inhalatory process, the more deeply particles of all sizes tend to be driven into the lung. Once deposited, materials must be in solution before they can be absorbed. Hence, materials in an aerosolized solution will be more readily absorbed than materials that are delivered as solid (e.g. dusts) particles. Solid materials must be able to go into solution in situ in order to be absorbed. Particle size influences dissolution rate. Large particles dissolve more slowly (for any given material) than small particles due to the differences in surface area. Once in solution, the same laws of passive diffusion apply to materials in the lung as apply to material in the GI tract. The large surface area and the rich blood flow at the alveoli make for ideal conditions for rapid absorption into the systemic circulation. Absorption across the mucosa lining the upper airways is less rapid. Materials that do not dissolve are ingested by pulmonary macrophages and either broken down there or moved out of the lungs by the upward movement of the bronchociliary tree. For gases and vapors, the amount absorbed is highly dependent on the partial pressure of the gas and the solubility of the gas in blood. Let us take the simple case of a gas that is not metabolized and is excreted by exhalation (e.g., an anesthetic gas or a Halon type fire extinguishing agent). At any given concentration (or partial pressure) in the atmosphere, the concentration in the blood will reach a steady state in the blood. Accordingly, prolonged exposure does not lead to continual build-up. At equilibrium, the concentration in the blood is depicted by the formula (also known as the Ostwald coefficient) Xb/Xa = S, where Xb is the concentration in the blood and Xa is the concentration in the inspired air. Thus, if one knows the S for a given chemical and the target concentration for a given exposure, one can predict what the resulting concentration may be at equilibrium. Additionally, the lower

Toxicokinetics in Biomaterial and Device Safety Evaluation

259

the S value (i.e., the lower the solubility in blood) the more rapidly the chemical will achieve equilibrium.

Absorption Across the Skin An aqueous carrier may be used for a variety of dermal products. In fact, carriers can be designed to limit the transportation of the penetration of the active ingredient (such as an insect repellent), if the desired effect is to keep the active on the surface of the skin. Once again, however, only those materials that are dissolved will be available for penetration across the skin to gain access to the systemic circulation. For almost all chemicals in or about to enter commerce, dermal penetration is a passive process. The relative thickness of the skin makes absorption (into the systemic circulation) slower than the absorption across the GI or pulmonary barriers. This is compounded by the fact that the stratum corneum function is to be impervious to the environment. One of the skin’s major functions is protection from infection. Once a chemical penetrates into the dermis, it may partition into the subcutaneous fat. Essentially, absorption across the skin is a two step process with the first being penetration and deposition into the skin and the second being release from the skin into the systemic circulation. The pattern of blood levels obtained via dermal penetration is generally one with a delayed absorption, slow build-up to more of a plateau than a peak. Blood levels of chemicals absorbed via the dermal route are generally low. Given the overwhelming influence of the physical properties of skin in determining bioavailabilities via the dermal route, assessment of dermal penetration is one area in metabolism and toxicology where the use of in vitro methods can be effectively used to predict in vivo results and to screen chemicals. Apparatus and equipment exist that one can use to maintain sections of skin (obtained from euthanized animals or from human cadavers or surgical discard) for such experiments (for examples, see 6). These apparatus are set up to maintain the metabolic integrity of the skin sample between two reservoirs: the one on the stratum corneum side, called the application reservoir, and the one on the subcutaneous side, called the receptor reservoir. One simply places radiolabeled test material in the application reservoir and collects samples at various time points from the receptor fluid. The rate of penetration can be presented by the traditional kinetic formulae to obtain a penetration rate constant. Given that exposed surface area also plays a role in the amount of material absorbed, the concept of flux is also important. The three major considerations in determining the quantity of material that is absorbed into the skin, and eventually released into the systemic circulation, is primarily dependent upon three factors: the surface area exposed, the volume of material applied, and the concentration of the material applied and the nature of the vehicle. Surface area—all things being equal, it is clear that the greater the surface exposed, the higher the achieved internal dose. Volume—the volume of material will obviously play a role in total dose, but it is not as straightforward as the relationship to surface area. Theoretically, the maximum absorption is obtained when the material is spread as thin and uniform as possible; piling material on so that it is literally rolling off the animal serves no practical purpose. In fact, it is not a sound practice when dealing with an in vivo animal experiment as it makes it more likely for the material to be available for oral ingestion.

260

Safety Evaluation of Medical Devices

Parameters Controlling Absorption The absorption of a chemical into the skin is a function of the nature of the molecule, the behavior of the vehicle, and the status of the skin. Three major variables account for differences in the rate of absorption or flux of different topical chemicals or of the same molecule in different vehicles: the concentration of the molecule in the vehicle, the partition coefficient of chemical between the stratum corneum and the vehicle, and the diffusion coefficient of molecule in the stratum corneum. The rate of diffusion is proportional to the concentration of molecule in the vehicle. The relationship is linear only at low molecule concentrations and only applies to soluble molecule in the vehicle. The latter factor may explain the variable therapeutic effects of different formulations of the same drug molecule. The partition coefficient is a measure of the molecule’s ability to escape from the vehicle and is defined as the equilibrium solubility of molecule in the surface of the stratum corneum relative to its solubility in the vehicle. Increased lipid solubility favors penetration of molecule through the skin by increasing the solubility in the relatively lipophilic stratum corneum. The diffusion coefficient indicates the extent to which the matrix of the barrier restricts the mobility of the molecule. Increases in molecular size of the molecule will increase frictional resistance and decrease the diffusion coefficient (Bronaugh, 1998); molecules over 1000 daltons usually will not be absorbed easily into normal adult skin. Finally, intact stratum corneum is an excellent barrier, but in disease states the resistance to absorption is rapidly lost and absorption can be facilitated.

Distribution Once the chemical gains access to the body, it is carried by the bloodstream and distributed to the different organs. The preferential organ of deposition is determined by a variety of

Table 19.2

factors: the two most important are blood flow to the organ and the affinity of the chemical for that organ. Affinity is governed by two general characteristics. Firstly, the product may be designed to have a specific affinity for a specific molecular entity in a target cell. For example, an anticholinesterase insecticide will tend to accumulate in the cells that have the highest concentration of cholinesterase. Secondly, the product may have a nonspecific or general chemical attraction for a specific cell type. The more highly lipophilic a chemical, the more likely it is to distribute and remain in adipose tissue. Blood flow will also have a major impact on distribution, as chemicals will be distributed more readily to those organs that are more highly perfused. A highly lipophilic chemical may first be deposited in the brain due to the fact that it is richly perfused, and then be distributed to body fat with time. Once a material is absorbed, distribution of a compound in most toxicology studies is usually of limited interest. Some factors which can serve to alter distribution are listed in Table 19.2. For most drugs, the rate of disposition or loss from the biological system is independent of rate and input, once the agent is absorbed. Disposition is defined as what happens to the active molecule after it reaches a site in the blood circulation where concentration measurements can be made (the systemic circulations, generally). Although disposition processes may be independent of input, the inverse is not necessarily true because disposition can markedly affect the extent of availability. Agents absorbed from the stomach and the intestine must first pass through the liver before reaching the general circulation (Figure 19.1). Thus, if a compound is metabolized in the liver or excreted in bile, some of the active molecule absorbed from the gastrointestinal tract will be inactivated by hepatic processes before it can reach the systemic circulation and be distributed to its sites of action. If the metabolizing or biliary excreting capacity of the liver is

Selected Factors That May Affect Chemical Distribution to Various Tissues

Factors relating to the chemical and its administration Degree of binding of chemical to plasma proteins (i.e., agent affinity for proteins) and tissues Chelation to calcium, which is deposited in growing bones and teeth (e.g., tetracyclines in young children) Whether the chemical distributes evenly throughout the body (one compartment model) or differentially between different compartments (two or more compartment model) Ability of chemical to cross the blood–brain barrier Diffusion of chemical into the tissues or organs and degree of binding to receptors that are and are not responsible for the drug’s beneficial effects Quantity of chemical given Route of administration/exposure Partition coefficients (nonpolar chemicals are distributed more readily to fat tissues than are polar chemicals) Interactions with other chemicals that may occupy receptors and prevent the drug from attaching to the receptor, inhibit active transport, or otherwise interfere with a drug’s activity Molecular weight of the chemical Factors relating to the test subject Body size Fat content (e.g., obesity affects the distribution of drugs that are highly soluble in fats) Permeability of membranes Active transport for chemicals carried across cell membranes by active processes Amount of proteins in blood, especially albumin Pathology or altered homeostasis that affects any of the other factors (e.g., cardiac failure and renal failure) The presence of competitive binding substances (e.g., specific receptor sites in tissues bind drugs) pH of blood and body tissues pH of urinea Blood flow to various tissues or organs (e.g., well-perfused organs usually tend to accumulate more chemical than less well-perfused organs) a

The pH of urine is usually more important than the pH of blood.

Chapter 19:

great, the effect on the extent of availability will be substantial. Thus, if the hepatic blood clearance for the chemical is large, relative to hepatic blood flow, the extent of availability for this chemical will be low when it is given by a route that yields first-pass metabolic effects. Likewise, metabolism is generally of only limited concern in most acute studies. There are some special cases, however, in which metabolic considerations must be factored in seeking to understand differences between routes and the effects which may be seen. The first special case is parenteral routes, where the systemic circulation presents a peak level of the moiety of interest to the body at one time, tempered only by the results of a single pass through the liver. The second special case arises from inhalation exposures. Because of the arrangements of the circulatory system, inhaled compounds enter the full range of systemic circulation without any “first pass” metabolism by the liver. Keberle et al. (1971) and O’Reilly (1972) have published reviews of absorption, distribution, and metabolism that are relevant.

Protein Binding The degree to which a chemical binds to plasma proteins will highly influence its distribution. Albumin, the most prominent of the many proteins found in mammalian plasma, carries both positive and negative charges with which a polar compound can associate by electrostatic attraction. As with all such reactions it can be described by the following equations. The more avidly bound the material, the less will be distributed to surrounding fluids as part of a solution and only that portion that is free in solution will be available for diffusion into the tissues.

Water Solubility The solubility of a chemical has a direct bearing on its distribution. Recall that only molecules that are in solution will be available for absorption. As mentioned above, only that portion that is free in solution will be available for diffusion into the tissues. Hence, the more material that is in solution the more that will be available for diffusion.

Volume of Distribution If one takes the dose administered (mg) and divide it by the plasma concentration of the test material (mg/mL), the result is a volume number. Dose = Volume Concentration One can take this process a step further and extrapolate back from a plasma time curve to the y axis. This is theoretically the plasma concentration (C0) that would occur if, upon being administered, the material is instantly distributed throughout the body. The volume number obtained with the above equation becomes: Dose C0

= VD

where VD represents the apparent volume of distribution, a proportionality constant that reflects the relation of the concentration of a xenobiotic in plasma to the total amount

Toxicokinetics in Biomaterial and Device Safety Evaluation

Table 19.3

261

Volume and Half-Life of Body Water in Selected Species

Species

Sex

Mouse Rat Rabbit Dog Cynomolgus monkey Rhesus monkey Man

F M F M M M M,F

Exchangeable body water (% of body weight) 58.5 59.6 58.4 66.0 61.6 61.6 55.3

Half-life (days) 1.13 2.53 3.87 5.14 7.80 7.80 9.46

of the entity in the body. Materials that are avidly bound to plasma proteins will have a high, while materials that are avidly taken by the tissues (deposit fat, for example) will have a low volume of distribution. The VD is a parameter that is simple to calculate yet gives one important information about the distribution of the chemical under investigation. The available volumes and masses for distribution vary from species to species, as summarized in Tables 19.3 and 19.4.

Metabolism/Biotransformation Metabolism describes the process by which chemicals are changed by the body. In fact, very few foreign chemicals that come to enter the body are excreted unchanged. Most are chemically modified. In general, metabolism results in chemicals that are more polar and water soluble, and more easily excreted. Examples of more common metabolic conversions are shown in Table 19.5. In general, the vast majority of lipophilic chemicals are first oxidized via the cytochrome P-450 dependent mixed function oxidase system of the liver. This is the process classically called Phase I metabolism. Cytochrome P-450 exists as a family of isozymes (the CYP gene superfamily) with varying but overlapping substrate affinity and responses to different inducing agents. For a review of the molecular biology of the CYP gene superfamily the reader is referred to Meyer (1994). Induction is the process whereby exposure to a chemical leads to increased activity of the MMFO due to an increase in cytochrome P-450. The isoenzymes induced by a variety of different chemicals are given in Table 19.6. In a practical sense, a chemical can induce its own metabolism. Hence, repeated dosing with a chemical may lead to lower blood levels at the end, for example, of a thirteen week study than at the beginning. There could also be alterations in the spectrum of metabolites produced, such that an agent could become more, or less, toxic with repeated dosing depending on the nature of the metabolites. It is not unusual during a Table 19.4

Typical Organ Weights in Adult Laboratory Animals Percent of body weight

Organ

Rat

Mouse

Dog

Rabbit

Monkey

Liver Kidney Heart Spleen Brain Adrenals Lung

3.5 0.8 0.4 0.3 0.5 0.02 0.6

6 1.6 0.4 0.5 0.6 0.01 0.6

3.5 0.5 0.8 0.3 0.8 0.01 1

3 0.8 0.3 0.04 0.4 0.02 0.6

2.5 0.5 0.4 0.1 3 0.03 0.7

262

Safety Evaluation of Medical Devices

Table 19.5

Summary of Prominent Phase I Biotransformation Reactions

Reaction

Enzyme

Location

Example/Comments

Hydrolysis

Carboxylesterase Peptidase

Ubiquitous Blood, lysomes

Epoxide hydroplase

Microsomes, cytosol

Vinyl acetate to acetate and acetaldehyde Amino-, carboxy- and endo peptidase which cleave peptides at specific amino acid linkages Conversion of styrene 7,8 epoxide to styrene 7,8 glycol

Reductions

azo and nitro reduction Carbonyl reductase Disulfide reduction Sulfoxide reduction Quinone reduction Reductive dehalogenation

Gut microflora Cytosol Cytosol Cytosol Cytosol, microsomes Microsomes

Sequential conversion of nitrobenzene to aniline Conversion of haloperidol to reduced haloperidol (a secondary alcohol) Glutathine dependent reduction of disulfiram to diethyldithiocarbamate Thioredoxin dependent of sulindac to sulindac sulfide DT diaphorase reduction of menadione to hydroquinone Conversion of pentabromoethane to tetra bromoethane (releasing free bromide ion)

Oxidation

Alcohol dehydrogenase

Cytosol

Aldehyde dehydrogenase Aldehyde oxidase Xanthene oxidase Monamine oxidase Diamine oxidase

Mitochondria/cytosol Liver cytosol Cytosol Mitochondria Cytosol

Prostaglandin oxidase

Microsomes

Flavin-monooxygenase

Microsomes

Conversion of ethaol to acetaldehyde (DAD/DADH dependent reversible reaction) Conversion of acetaldehyde to acetate FAD dependent metalloenzyme, oxidation of benzaldehyde to benzoic acid Oxidation of purine derivative, conversion of allopurinol to alloxanthene FAD dependent oxidative deamination of monoamines, e.g. primaquine Pyridoxal dependent, copper containing enzyme, conversion of allylamine to acrolein Cooxidation reaction, can activate chemical in tissues low in cytochrome P-450, e.g. nephrotoxicity of acetaminophen, oxidation of phenylbutazone FAD dependent oxidation of nucleophilic nitrogen, sulfur and phosphorus heteroatoms, e.g., conversion of nicotine to nicotine 1′-N-oxide, cimetidine to cimetidine S-oxide

Cytochrome P-450

Microsomes

subchronic or chronic toxicity test for tolerance to occur. There may be signs of toxicity early in the study but even with continued daily dosing, the signs abate. This phenomenon, particularly in rodents, is frequently due to microsomal induction, whereby the chemical has induced its own metabolism, and more rapid clearance of the parent chemical occurs. After the chemical has been metabolically oxidized, it can in fact be further metabolized. In fact, it is possible for the metabolites to also be substrates of the MMFO and to be metabolized themselves. The route of metabolic activation of the classic carcinogen benzo[a]pyrene is due to such a mechanism. The biology of these reactive intermediates has been extensively studied. Glutathione is among the most common organic intracellular chemicals in all mammalian species, being present at a concentration of up to 10 mm (need to check) and glutathione S-transferase is very active. Glutathione is a tripeptide (glutamine-cysteine glycyne). The sulfhydryl group of cysteine is the business end of the molecule where the reaction with the nucleophilic reactive intermediate takes place. After that, the glutathione conjugate is further metabolized to a cytinyl–acetyl moiety. These moieties are called mercapturic acids, and are generally found in the urine. The relative predominance of mercapturic acid over other metabolites may be considered a rough indication of how “reactive” the intermediates may have been. Teleological, it is tempting to speculate that it is a very well designed protective mechanism. So long as intracellular glutathione concentrations remain above a critical level the destructive actions of active metabolites can be held in check. Thus, a small dose of a chemical (bromo-benzene is a good example) may cause no liver damage while a large dose may. This is also a good example of one of the aspects of toxicokinetics versus pharmacokinetics where a high dose

of a chemical will become toxic due to saturation of a detoxification pathway. The glutathione S-transferase pathway is sometimes in biochemical competition with the epoxide hydratase pathway, in that both deactivate intermediates of the MMFO. Epoxide hydratase is a microsomal enzyme that acts specifically to deactivate epoxide intermediates, by the addition of water across the C–O bond to form a diol. As a very broad generality, the glutathione S-transferase pathway tends to be more prominent in rodents, while the epoxide hydratase pathway tends to be more dominant in nonrodents. The hydroxyl or diol containing metabolites of the MMFO can be further metabolized by so-called phase II (synthetic) metabolism, whereby they are conjugated to/from glucuronides and/or sulfates (so called etherial sulfates). Amines can also be substrates. The net effect of phase II reactions is to create a more polar molecule that is more readily excretable. While there are species differences, glucuronides are actively transported and excreted in the bile into the GI tract. Sulfates are excreted more predominantly in the urine. Both glucuronides and sulfates, however, can be found in both the urine and the feces. Like the MMFO pathway, glutathione 3-transferase, UDP-glucuronyl transferase and eopoxide hydratase are inducible, i.e. treatment with exogenous chemicals will increase the amount of enzyme protein present. Outside of the MMFO mediated (Phase I) reactions there are a few other major reactions that are worthy of note. The two major ones involve ester hydrolysis and alcohol and aldehyde dehydrogenases. All mammalian species have an extensive ability to hydrolyze the ester bond. The products of the reactions then can go on to be further metabolized. In the pharmaceutical industry, this property has been utilized to synthesize pro-drugs; i.e., chemicals that have desirable pharmaceutical properties (generally increased water solubility)

Chapter 19:

Table 19.6

Examples of Xenobiotics Activated by Human P450

CYP1A1 Benzo[a]pyrene and other polycyclic aromatic hydrocarbons CYP1A2 Acetaminophen 2-Acetylaminofluorene 4-Aminobiphenyl 2-Aminofluorene 2-Naphthylamine NNK* Amino acid pyrolysis products (DiMeQx, MelQ, MelQx, Glu P-1, Glu P-2, IQ, PhlP, Trp P-1, Trp P-2) CYP2A6 N-nitrosodiethylamine NNK* CYP2B6 6-Aminochrysene Cyclophosphamide Ifosphamine CYP2C8, 9, 18, 19 None known CYP2D6 NNK* CYP2E1 Acetaminophen Acrylonitrile Benzene Carbon tetrachloride Chloroform Dichloromethane 1,2-Dichloropropane Ethylene dibromide Ethylene dichloride Ethyl carbamate N-Nitrosodimethylamine Styrene Trichloroethylene Vinyl chloride CYP3A4 Acetaminophen Aflatoxin B1 and G1 6-Aminochrysene Benzo[a]pyrene 7,8-dihydrodiol Cyclophosphamide Ifosphamide 1-Nitropyrene Sterigmatocystin Senecionine Tris(2,3-dibromopropyl) phosphate CYP4A9/11 None known

that are not converted to their active moiety until hydrolyzed in the body. The activity of alcohol dehydrogenase is one with which we should all be familiar. It oxidizes alcohols to aldehydes. The aldehydes produced by this reaction can go on to be further metabolized to a carbocylic acid, if they are not sterically hindered. Sidechain constituents of aromatic compounds can also be a substrate for this reaction sequence, producing side chain carboxylates. The oxidation of alcohols to aldehydes can also be a form of metabolic activation as aldehydes can have potent physiological actions. Fortunately, aldehyde dehydrogenase has a very high activity when compared to alcohol dehydrogenase, so that the aldehydes do not accumulate. Inhibition of aldehyde

Toxicokinetics in Biomaterial and Device Safety Evaluation

263

dehydrogenase by disulfiram (Antabuse) leads to the accumulation of acetaldehyde, causing nausea, dizziness and flushing. Like disulfiram, some pesticides contain dithiocarbamates and have the potential of causing this type of reaction. Hopefully, this little description of the major metabolic pathways has given one some appreciation of the richness of the processes. The different sites of oxidation, the possibility of additional oxidative metabolism of metabolites, differences in Phase II reactions all lead to a multiplicity of possible metabolites. Over 100 different metabolites of the human pharmaceutical chlorpromazine have been isolated and identified. When analyzed by HPLC, for example, the parent chemical and the different (detectable) metabolites will form a pattern of different peaks. This is referred to as the metabolic fingerprint or profile of a chemical. Different species will have different profiles. Ideally, in doing a risk assessment, one would like to know the similarity in this pattern between the animals used in the toxicology studies and that produced by human beings. This is only infrequently available for most non-pharmaceutical products, as pesticides (for example) are rarely given intentionally to human subjects for the purposes of study. The technology now exists, however, to address this potential problem. Cell lines with human cytochrome P-450 have been developed that can provide some indication of the similarities of human metabolism of a chemical to that of experimental animals. At least they may be able to assist in identifying the major oxidative metabolite. For non-pharmaceutical products, if may be an unusual circumstance that would require one to identify potential human metabolites as part of a marketing application; however, it may be useful for one to know that the technology exists to do so. The processes of metabolic conversion are frequently involved in the mechanisms of toxicity and carcinogenicity.

Species Differences Species differences in metabolism are amongst the principal reasons that there are species differences in toxicity. Differences in cytochrome P450 is one of the most common reasons for differences in metabolism. For example, Monostory et al. (1997) recently published a paper comparing the metabolism of panomifene (a tamoxifen analog) in four different species. These data serve to address that the rates of metabolism in the non-human species was most rapid in the dog and slowest in the mouse. Thus, one should not a priori make any assumptions about which species will have the more rapid metabolism. Of the seven metabolites, only one was produced in all four species. Both the rat and the dog produced the two metabolites (M5 and M6) produced by human microsomes. So how does one decide which species best represents man? One needs to consider the chemical structure of the metabolites and the rates at which they are produced. In this particular case, M5 and M6 were relatively minor metabolites in the dog, which produced three other metabolites in larger proportion. The rat produced the same metabolites at a higher proportion, with fewer other metabolites than the dog. Thus, in this particular instance the rat, rather than the dog, was a better model. Likewise, Table 19.7 offers a comparison of excretion patterns between three species for a simple inorganic compound. A more thorough review on species differences in pharmacokinetics has been presented by Smith (1991) and in Gad and Chengelis (1992).

264

Safety Evaluation of Medical Devices

Table 19.7 Differences in the Disposition of 2,4-dichlorophenoxyacetic Acid Species

Sex

Urine

Feces

Rat

M F M F M F

31.2 16.5 12.7 26.8 4.9 33.9

2.7 1.1 2.8 6.7 2.5 14.5

Mouse Hamster

All animals dosed orally with radiolabeled 2.4-D, 200 mg/kg. Results are expressed as percent of 14C dose recovered. Urine was collected for 8 hours and feces for 24 hours.

carbonate (increases pH); the buffered plasma shows little change. 3. Xenobiotics may be secreted actively into the renal tubule against a concentration gradient by anion and cation carrier processes. These processes are saturable and of relatively low specificity; many basic or acidic compounds and their metabolites (especially conjugation products) are removed by them. Because the dissociation rate for the chemical–albumin complex is rapid, it is possible for highly protein-bound compounds to be almost completely cleared at a single passage through the kidney.

Feces Sex Related Differences in Rodents Not only are there differences in absorption, distribution, biotransformation and metabolism between species, there may also be differences between sexes within a species. Griffin et al. (1997), for example, has demonstrated sexrelated differences in the metabolism of 2,4-dichlorophenoxyacetic acid (Table 19.7). They noted that while there were differences between sexes, they tended to be quantitative (rates), not qualitative (metabolites). Differences between species were greater than sex-related differences. With regard to sex-related differences, it is noteworthy that males do not always have the higher rates, as Griffien et al. (1997) have shown; in hamsters, the female metabolizes 2,4-D more rapidly than males. In general, male rats tend to have higher activity than female rats, especially with regard to CYP dependent activity. In the case of 2,4-D, the only urinary metabolite is 2,4-D glucuronide, but the half life of 2,4-D was 138 min in males and 382 in females.

Excretion Excretion encompasses the process by which chemicals or their metabolites are transported out of the body. There are three possible major routes of excretion, and a handful of minor ones. The major routes of excretion for chemicals, and in particular their metabolites, are:

Urine The kidneys filter the entire cardiac output multiple times each day, and thus provide a large opportunity for the removal of chemicals from the bloodstream. How much of a xenobiotic is actually excreted is dependent on three factors or processes. 1. The glomerular membrane has pores of 70–80 Å; and under the positive hydrostatic conditions in the glomerulus, all molecules smaller than about 20,000 Da are filtered. Proteins and protein-bound compounds thus remain in the plasma, and about 20% of the nonbound entity is carried with 20% of the plasma water into the glomular filtrate. 2. Because the glomerular filtrate contains many important body constituents (e.g., glucose), there are specific active uptake processes for them. Also, lipid-soluble chemicals diffuse back from the tubule into the blood, especially as the urine becomes more concentrated because of water reabsorption. The pH of the urine is generally lower than that of the plasma, and therefore pH partitioning tends to increase the reabsorption of weak acids. The pH of the urine can be altered appreciably by treatment with ammonium chloride (decreases pH) or sodium

The most important mechanism allowing circulating foreign compounds to enter the gut is in the bile. The biological aspects of this mechanism have been reviewed (80), and certain pertinent points have emerged. The bile may be regarded as a complementary pathway to the urine, with small molecules being eliminated by the kidney and large molecules in the bile. Thus the bile becomes the principal excretory route for many drug conjugates. Species differences exist in the molecular weight requirement for significant biliary excretion, which has been estimated as 325 ± 50 in the rat, 440 ± 50 in the guinea pig, and 475 ± 50 in the rabbit. In the rat, small molecules (less than 350 Da) are not eliminated in the bile or large molecules (more than 450 Da) in the urine, even if the principal excretory mechanism is blocked by ligation of the renal pedicles or bile duct, respectively. Compounds of intermediate molecular weight (350–450 Da) are excreted by both routes, and ligation of one pathway results in increased use of the other. Foreign compounds may also enter the gut by direct diffusion or secretion across the gut wall, elimination in the saliva, pH partitioning of bases into the low pH of the stomach, and elimination in the pancreatic juice.

Expired Air Volatile compounds or metabolites can be extensively excreted by passage across pulmonary membranes into the airspace of the lungs, then expulsion from the lungs in expired air. Minor routes for excretion can include tears, saliva, sweat, exfoliated keratinocytes, hair, and nasal discharge. These are of concern or significance only in rare cases. Accordingly, quantitation of excretion typically requires collection of urine, feces (and occasionally expired air) over a period of time.

Pharmacokinetics The interplay of the processes of absorption, distribution, metabolism and excretion result in changes in concentration of the test chemical in different organs with time. With regard to the practical concerns of monitoring human exposure, the organ of interest is the blood. Blood can be considered a central compartment. Determining the concentration of the chemical in plasma gives one an assessment of exposure. Mathematical formulas are used to quantitatively describe this exposure.

Physiologically-based Pharmacokinetic Modeling Pharmacokinetic parameters are descriptive in nature. They quantitatively describe the manner in which a test material is absorbed and excreted, such that a specific blood or tissue

Chapter 19:

level is achieved or maintained. In the past, experiments had to be done by every route of administration to gather the data appropriate for describing the pharmacokinetic behavior of a chemical administered by different routes. The development of more sophisticated and readily accessible computers has led to the development of a different approach, that of pharmacokinetic modeling. In this computerized model, different compartments are represented as shown in boxes and the movement of the material in and out of the compartments is defined by the rate constants. These can be determined either in vivo or in vitro. Other physiological parameters are brought into play as well, such as octanol/water partition coefficient, blood flow through an organ, respiration rate (for the inhalation route of exposure), rate of microsomal metabolism, etc.

Toxicokinetics in Biomaterial and Device Safety Evaluation

265

During the synthesis of radiolabeled compounds, the label is usually introduced as part of the molecular skeleton in a metabolically stable and, with tritium, nonexchangeable position. The in vivo stability of 14C labels is often reflected by the extent of [14C] carbon dioxide formation. The biologic stability of 3H labels can be estimated by the extent of tritiated water formation. The tritiated water concentration (dpm/mL) in urine samples collected during a designated time interval after dosing, assumedly after equilibrium is reached between urine and the body water pool is determined. This value is extrapolated from the midpoint of the collection interval to zero time, based on the known half-life of tritiated water in the given species. The percentage of the radioactive dose that is transformed to tritiated water (% 3H2O) can be calculated using the following equation: H 2 O concentration at zero time × exchangeable body water volume × 100% % H 2O = radioactivity dose 3

LABORATORY METHODS

3

The actual means by which pharmacokinetic information is collected is through the conduct of one or more of specific studies, employing a wide range of available analytical techniques. Administered therapeutic molecules can be identified and quantified in relevant samples collected in accordance with carefully designed and executed protocols.

Radiochemical Methods The massive expansion of our understanding of toxicokinetics since the late 1970s is to a large degree a reflection of the wide use of radioactive isotopes as tracers of chemical and biological processes. Appropriately radiolabeled test compounds are commonly used in toxicokinetic studies, providing a simple means of following the administered dose in the body. This is particularly important when specific analytical methods are unavailable or too insensitive. The use of total radioactivity measurements allows an estimation of the total exposure to drug-related material and facilitates the achievement of material balance. The most commonly used radionuclides in drug metabolism and disposition studies are carbon-14 (14C) and tritium ( 3H), both of which are referred to as beta emitters. Since these beta-emitting isotopes have relatively long halflives, their radioactive decay during an experiment is insignificant. Additionally, they provide sufficient emission energy for measurement and are relatively safe to use, as indicated by the data in Table 19.8. Although individual beta particles can have any energy up to the maximum, Emax, the basic quantity in determining the energy imparted to tissues by beta emitters is the average energy, Eβ. The range is the maximum thickness the beta particles can penetrate. Beta particles present virtually no hazard when they originate outside the body (Shapiro, 1981). Table 19.8 Properties of Primary Radioisotopes Employed in Pharmacokinetics Property Half-life (yr) Maximum beta energy (MeV) Average beta energy (MeV) Range in air (mm) Range in unit density material (mm)

3

H

51

Cr

12.3 0.0186

27.8 (d) 0.752

0.006

0.049

6 0.0052

300 0.29

14

C

5730 0.156

125

I

13 (d) 2.150

Values for the exchangeable body water content as well as the half-life of tritiated water in some mammalian species that can be applied to the above equation were shown earlier in Table 19.3. If the molecule is likely to or is known to fragment into two major portions, it may be desirable to monitor both fragments by differential labeling (3H and 14C). The chemical and radiochemical purity of the labeled compound must be ascertained prior to use. In practice a value of 95% or greater is usually acceptable. The desired specific activity of the administered radioactive compound depends on the dose to be used as well as the species studied. Doses of 14C on the order of 5 µCi/kg for the dog and 20 µCi/kg for the rat have been found adequate in most studies, while doses of 3H are usually two to three times higher owing to lower counting efficiency of this isotope. Liquid scintillation counting is the most popular technique for the detection and measurement of radioactivity. In order to count a liquid specimen such as plasma, urine, or digested blood or tissues directly in a liquid-scintillation spectrometer, an aliquot of the specimen is first mixed with a liquid scintillant. Aliquots of blood, feces, or tissue homogenates are air-dried and ash-free filter papers and combusted in a sample oxidizer provided with an appropriate absorption medium and a liquid scintillant prior to counting. The liquid scintillant plays the role of an energy transducer, converting energy from nuclear decay into light. The light generates electrical signal pulses which are analyzed according to their timing and amplitude, and are subsequently recorded as a count rate, e.g., counts per minute (cpm). Based on the counting efficiency of the radionuclide used, the count rate is then converted to the rate of disintegration, e.g., disintegrations per minute (dpm), which is a representation of the amount of radioactivity present in the sample.

Immunoassay Methods Radioimmunoassay (RIA) allows measurement of biologically active materials which are not detectable by traditional cold chemistry techniques. RIAs can be used to measure molecules that cannot be radiolabeled to detectable levels in vivo. They also are used for molecules unable to fix complement when bound to antibodies, or they can be used to identify cross-reacting antigens that compete and bind with the antibody.

266

Safety Evaluation of Medical Devices

Competitive inhibition of radiolabeled hormone antibody binding by unlabeled hormone (either as a standard or an unknown mixture) is the principle of most RIAs. A standard curve for measuring antigen (hormone) binding to antibody is constructed by placing known amounts of radiolabeled antigen and the antibody into a set of test tubes. Varying amounts of unlabeled antigen are added to the test tubes. Antigen–antibody complexes are separated from the antigen and the amount of radioactivity from each sample is measured to detect how much unlabeled antigen is bound to the antibody. Smaller amounts of radiolabeled antigen–antibody complexes are present in the fractions containing higher amounts of unlabeled antigen. A standard curve must be constructed to correlate the percentage of radiolabeled antigen bound with the concentration of unlabeled antigen present. Two methods are commonly employed in RIAs to separate antigen–antibody complexes. The first, the doubleantibody technique, precipitates antigen–antibody complexes out of solution by utilizing a second antibody, which binds to the first antibody. The second most commonly used method is the dextran-coated activated charcoal technique. Addition of dextran-coated activated charcoal to the sample followed immediately by centrifugation absorbs free antigen and leaves antigen–antibody complexes in the supernatant fraction. This technique works best when the molecular weight of the antigen is 30 kilodaltons (kDa) or less. Also, sufficient carrier protein must be present to prevent adsorption of unbound antibody. Once a standard curve has been constructed, the RIA can determine the concentration of hormone in a sample (usually plasma or urine). The values of hormone levels are usually accurate using the RIA, but certain factors (e.g., pH or ionic strength) can affect antigen binding to the antibody. Thus similar conditions must be used for the standard and the sample. Problems of RIAs include lack of specificity. This problem is usually due to nonspecific cross-reactivity of the antibody. RIA represents an analytical approach of great sensitivity. Unlike assays that often require large amounts of tissue (or blood), the greater sensitivity of the RIAs or monoclonal antibody techniques can be achieved using small samples of biological fluids. Some of these RIA methodologies are more useful than others and to some extent depend on the degree of hormonal cross-reactions or, in the case of monoclonal antibody methods, their degree of sensitivity. Enzyme-linked immunosorbent assay (ELISA) is comparable to the immunoradiometric assay except that an enzyme tag is attached to the antibody instead of a radioactive label. ELISAs have the advantage of no radioactive materials and produce an end product that can be assessed with a spectrophotometer. The molecule of interest is bound to the enzyme-labeled antibody, and the excess antibody is removed for immunoradiometric assays. After excess antibody has been removed or the second antibody containing the enzyme has been added (two-site assay), the substrate and cofactors necessary are added in order to visualize and record enzyme activity. The level of molecule of interest present is directly related to the level of enzymatic activity. The sensitivity of the ELISAs can be enhanced by increasing the incubation time for producing substrate. Immunoradiometric assays (IRMAs) are like RIAs in that a radiolabeled substance is used in an antibody–antigen reaction, except that the radioactive label is attached to the antibody instead of the hormone. Furthermore, excess of antibody, rather than limited quantity, is present in the assay. All the unknown antigen becomes bound in an IRMA rather than just a portion, as in a RIA; IRMAs are more sensitive.

In the one-site assay, the excess antibody that is not bound to the sample is removed by addition of a precipitating binder. In a two-site assay, a molecule with at least two antibodybinding sites is adsorbed onto a solid phase, to which one of the antibodies is attached. After binding to this antibody is completed, a second antibody labeled with 125I is added to the assay. This antibody reacts with the second antibodybinding site to form a “sandwich,” composed to antibodyhormone-labeled antibody. The amount of hormone present is proportional to the amount of radioactivity measured in the assay. With enzyme-multiplied immunoassay technique (EMIT) assays, enzyme tags are used instead of radiolabels. The antibody binding alters the enzyme characteristics, allowing for measurement of target molecules without separating the bound and free components (i.e., homogeneous assay). The enzyme is attached to the molecule being tested. This enzyme-labeled antigen is incubated with the sample and with antibody to the molecule. Binding of the antibody to the enzyme-linked molecule either physically blocks the active site of the enzyme or changes the protein conformation so that the enzyme is no longer active. After antibody binding occurs, the enzyme substrate and cofactor are added, and enzyme activity is measured. If the sample contains subject molecules, it will compete with enzyme-linked molecules for antibody binding, enzyme will not be blocked by anti-body, and more enzyme activity will be measurable. Most chemical entities can now be assessed using monoclonal antibody (MAb) techniques. It is possible to produce antisera containing a variety of polyclonal antibodies that recognize and bind many parts of the molecule. Polyclonal antisera can create some nonspecificity problems such as cross-reactivity and variation in binding affinity. Therefore it is oftentimes desirable to produce a group of antibodies that selectively bind to a specific region of the molecule (i.e., antigenic determinant). In the past, investigators produced antisera to antigenic determinants of the molecule by cleaving the molecule and immunizing an animal with the fragment of the hormone containing the antigenic determinant of interest. This approach solved some problems with cross-reactivity of antisera with other similar antigenic determinants, but problems were still associated with the heterogeneous collection of antibodies found in polyclonal antisera. The production of MAbs offers investigators a homogenous collection of antibodies that could bind selectively to a specific antigenic determinant with the same affinity. In addition to protein isolation and diagnostic techniques, MAbs have contributed greatly to RIAs. While MAbs offer a highly sensitive, specific method for detecting antigen, sometimes increasing MAb specificity compromises affinity of the antibody for the antigen. In addition, there is usually decreased complement fixation, and costs are usually high for preparing and maintaining hybridomas that produce MAbs (Table 19.9). The monoclonal antibody techniques provide a means of producing a specific antibody for binding antigen. This technique is useful for studying protein structure relations (or alterations) and has been used for devising specific RIAs.

Sampling Methods and Intervals Blood Since blood (plasma and serum) is the most easily accessible body compartment, the blood concentration profile is most commonly used to describe the time course of drug

Chapter 19:

Table 19.9 Advantages and Disadvantages of Monoclonal Antibodies Compared to Polyclonal Antisera Advantages

Disadvantages

Sensitivity Quantities available Immunologically defined Detection of neoantigens on cell membrane

Overly specific Decreased affinity Diminished complement fixation Labor intensive; high cost

disposition in the animal. With the development of sensitive analytical methods that require small volumes (100–200 µL) of blood, ADME data from individual rats can be obtained by serial sample collection. Numerous cannulation techniques have been utilized to facilitate repeated blood collection, but the animal preparation procedures are elaborate and tedious and are incompatible with prolonged sampling periods in studies involving a large number of animals. In contrast, noncannulation methods such as collection from the tail vein, orbital sinus, or jugular vein are most practical. Significant volumes of blood can be obtained from the intact rat by cardiac puncture, although this method can cause shock to the animal system and subsequent death. Blood collection from the tail vein is a simple and rapid nonsurgical method which does not require anesthesia. A relatively large number of serial samples can be obtained within a short period of time. However, this method is limited to relatively small sample volumes (7250 µL per sample). Although larger volumes can be obtained by placing the rat in a warming chamber, this procedure could significantly influence the disposition of the test compound and therefore is not recommended for routine studies. Blood collected from the cut tail has been shown to provide valid concentration data for numerous compounds. The rat is placed in a suitable restrainer with the tail hanging freely. The tail is immersed in a beaker of warm water (37–40∞C) for 1–2 min to increase the blood flow. Using surgical scissors or a scalpel, the tail is completely transected approximately 5 mm above the tip. The tail is then gently “milked” by sliding the fingers down the tail from its base. It should be noted that excessive “milking” could cause damage to the blood capillaries or increase the white cell count in the blood. A heparinized micropipet of desired capacity (25–250 µL) is held at a 30 to 45∞ downward angle in contact with the cut end of the tail. This allows blood to fill the micropipet by capillary action. Application of gentle pressure with a gauze pad for approximately 15 sec is sufficient to stop bleeding. A sufficient number of serial blood samples may be obtained to adequately describe the blood level profile of a compound. If plasma is required, the blood may be centrifuged after sealing one end of the filled micropipet and placing it in a padded centrifuge tube. The volume of plasma is determined by measuring the length of plasma as a fraction of the length of the micropipet, multiplied by the total capacity of the pipet. The tube is then broken at the plasma/red blood cell interface and the sample is expelled using a small bulb. If serum is needed, the blood should be collected without using anticoagulants in the sampling tube. Serial blood samples can also be collected from the orbital sinus, permitting rapid collection of larger (1–3 mL) samples.

Toxicokinetics in Biomaterial and Device Safety Evaluation

267

Excreta Excretion samples commonly collected from the rat include urine, feces, bile, and expirated air. By using properly designed cages and techniques, the samples can be completely collected so that the mass balance is readily determined. These samples also serve to elucidate the biotransformation characteristics of the compound. These samples can be easily collected through the use of suitable metabolism cages. Since rodents are coprophagic, the cage must be designed to prevent the animal from ingesting the feces as it is passed. Other main features of the cage should include the ability to effectively separate urine from feces with minimal cross-contamination, a feed and water system that prevents spillage and subsequent contamination of collected samples, and collection containers that can be easily removed without disturbing the animal. Also, the cage should be designed so that it can be easily disassembled for cleaning or autoclaving. Following dose administration, rats are placed in individual cages. The urine and feces that collect in containers are removed at predetermined intervals. The volume of urine and the weight of feces are measured. After the final collection, the cage is rinsed, normally with ethanol or water, to assure complete recovery of excreta. If the rats are also used for serial blood sampling, it is important that bleeding be performed inside the cage to avoid possible loss of urine or feces.

Bile The bile is the pathway through which an absorbed compound is excreted in the feces. In order to collect this sample, surgical manipulation of the animal is necessary.

Expired Air For 14C-labeled chemicals, the tracer carbon may be incorporated in vivo into carbon dioxide, a possible metabolic product. Therefore, when the position of the radiolabel indicates the potential for biological instability, a pilot study to collect expired air and monitor its radioactivity content should be conducted prior to initiating a full-scale study. Expired air studies should also be performed in situations where the radiolabel has been postulated to be stable but analyses of urine and feces from the toxicokinetic study fail to yield complete recovery (mass balance) of the dose. Following drug administration, the rat is placed in a special metabolism cage. Using a vacuum pump, a constant flow of room air (approx. 500 mL/min) is drawn through a drying column containing anhydrous calcium sulfate impregnated with a moisture indicator (cobalt chloride), and passed into a second column containing Ascarite® II, where it is rendered carbon dioxide free. The air is then drawn in through the top of the metabolism cage. Exhaled breath exiting the metabolism cage is passed through a carbon dioxide adsorption tower, where the expired 14CO2 is trapped in a solution such as a mixture of 2-ethoxyethanol and 2-aminoethanol (2:1). The trapping solution is collected, replaced with fresh solution, and assayed at designated times postdose so that the total amount of radioactivity expired as labeled carbon dioxide can be determined.

Milk The study of passage of a xenobiotic into milk serves to assess the potential risk to breast-fed infants in the absence of human data. The passage into milk can be estimated as the milk–plasma ratio of drug concentrations at each sampling

268

Safety Evaluation of Medical Devices

Table 19.10 Approximate Volumes of Pertinent Biological Fluids in Adult Laboratory Animals

Table 19.11 Blood Samples Required so that Certain Toxicokinetic Parameters can be Obtained and Calculated.

Fluid

Rat

Mouse

Dog

Rabbit

Parameter

Blood Sample Required

Information Obtained

Blood (ml/kg) Plasma (ml/kg) Urine (ml/kg/day) Bile (ml/kg/day)

75 40 60 90

75 45 50 100

70 40 30 12

60 30 60 120

Cmin (C24) CT Cmax (C peak) Tmax (T peak) AUC t1⁄2 Accumulation ratio

24-h T-h Several* Several* Several* Several* Several after first and repeated doses

Accumulation Proof of absorption Rate of absorption Rate of absorption Extent of absorption Various Extent of accumulation

Monkey 75 45 75 25

time or that of the AUC values. Approximately 30 rats in their first lactation are used. The litter size is adjusted to about 10 within 1–2 days following parturition. The test compound is administered to the mothers 8–10 days after parturition. The rats are then divided into groups for milk and blood collection at designated times postdose. All sucklings are removed from the mother rats several hours before milking. Oxytocin, 1 IU per rat, is given intramuscularly 10–15 min before each collection of milk to stimulate milk ejection. The usual yield of milk is about 1 mL from each rat. Blood is obtained immediately after milking. In order to minimize the number of animals used, the sucklings can be returned to the mother rat which can then be milked again 8–12 hours later. In all the fluid sampling techniques above, the limitations of availability should be kept in mind. Table 19.10 presents a summary of such availability for the principle model species. For topical exposures, determining absorption (into the skin and into the systemic circulation) requires a different set of techniques. For determining how much material is left, skin washing is required. There are two components to skin washing in the recovery of chemicals. The first component is the physical rubbing and removal from the skin surface. The second component is the surfactant action of soap and water. However, the addition of soap effects the partitioning. Some compounds may require multiple successive washing with soap and water applications for removal from skin. Skin tape stripping can be used to determine the concentration of chemical in the stratum corneum at the end of a short application period (30 min) and by linear extrapolation predicts the percutaneous absorption of that chemical for longer application periods. The chemical is applied to skin of animals or humans, and after a 30-minute skin contact application time, the stratum corneum is blotted and then removed by successive tape applications. The tapestrippings are assayed for chemical content. There is a linear relationship between this stratum corneum reservoir content and percutaneous absorption. The major advantages of this method are (1) the elimination of urinary and fecal excretion to determine absorption and (2) the applicability to nonradiolabeled determination of percutaneous absorption, because the skin strippings contain adequate chemical concentrations for nonlabeled assay methodology. Finally, a complete determination of the distribution and potential departing of a chemical and its metabolites requires some form of measurement or sampling of tissues/organs. Autoradiography provides a nonquantitative means of doing such, but quantitation requires actual collection and sampling of tissues. Table 19.6 provided guidance as to the relative percentage of total body mass that the organs constitute in the common model species.

Sampling Interval To be able to perform valid toxicokinetic analysis, it is not only necessary to properly collect samples of appropriate

*Several samples to define concentration-time profile.

biological fluids, but also to collect a sufficient number of samples at the current intervals. Both of these variables are determined by the nature of the answers sought. Useful parameters in toxicokinetic studies are Cmax, which is the peak plasma test compound concentration, Tmax, which is the time at which the peak plasma test compound concentration occurs, Cmin, which is the plasma test compound concentration immediately before the next dose is administered, AUC, which is the area under the plasma test compound 1 concentration–time curve during a dosage interval, and t ⁄ 2, which is the half-life for the decline of test compound concentrations in plasma. The samples required to obtain these parameters are shown in Table 19.11. Cmin requires one blood sample immediately before a dose is given and provides information on accumulation. If there is no accumulation in plasma, the test compound may not be detected in this sample. Several Cmin samples are required at intervals during the toxicity study to check whether accumulation is occurring. CT is a blood sample taken at a chosen time after dosing and provides proof of absorption, as required by the GLP regulations, but little else. Cmax requires several blood samples to be taken for its accurate definition as does Tmax; these two parameters provide information on rate of absorption. AUC also requires several blood samples to be taken so that it can be calculated; it provides information on extent 1 of absorption. t ⁄ 2, the half-life, requires several samples to be taken during the terminal decline phase of the test compound concentration–time curve; this parameter provides information on various aspects such as any change in the kinetics of the test compound during repeated doses or at different dose levels. Depending on the other parameters obtained, the accumulation ratio can be calculated from Cmin, Cmax and/or AUC when these are available after the first dose and after several doses to steady-state. Operational and metabolic considerations generally make urine sampling and assay of limited value for toxicokinetic purposes.

Study Type Metabolic and pharmacokinetic data from a rodent species and a non-rodent species (usually the dog) used for repeat dose safety assessments (14 days, 28 days, 90 days or six months) are recommended. If a dose dependency is observed metabolic and pharmacokinetic or toxicity studies with one species, the same range of doses should be used in metabolic and pharmacokinetic studies with other species. If human metabolism and pharmacokinetic data also are available, this information should be used to help select test species for the full range of toxicity tests, and may help to

Chapter 19:

justify using data from a particular species as a human surrogate in safety assessment and risk assessment. Metabolism and pharmacokinetic studies have greater relevance when conducted in both sexes of young adult animals of the same species and strain used for other toxicity tests with the test substance. The number of animals used in metabolism and pharmacokinetic studies would be sufficient to reliably estimate population variability. This usually means a separate (but parallel) set of groups of animals in rodent studies. A single set of intravenous and oral dosing results from adult animals, when combined with some in vitro kinetic results, may provide an adequate data set for the design and interpretation of short-term, subchronic and chronic toxicity studies. Studies in multiple species may clarify what appear to be contradictory findings in toxicity studies (i.e., equal mg/kg bw doses having less effect in one species that in another). If disposition and metabolite profiles are found to be similar, then differences in responses among species could more reliably be attributed to factors other than differences in metabolism. Studies of the pharmacokinetics and metabolism of a substance in neonatal and adolescent animals provide information about any changes in metabolism associated with tissue differentiation and development. Animals with fetuses of known gestational age should be used for determining the disposition of the test substance in the fetus. Dosage is by (to the maximum extent possible) the intended clinical route. An acute intravenous study can provide accurate rates of metabolism—without interference from intestinal flora— plus rats of renal and biliary elimination, if urine and bile are collected. This route also avoids the variability in delivered dose associated with oral absorption and ensures that the maximum amount of radiolabel is excreted in the urine or bile for purposes of detection. Once IV data and parameters are available, they can be used with plasma concentrations from limited oral studies to compute intestinal absorption via the ratio of Areas Under the (plasma and/or urine) Curves or via simulations of absorption with gastrointestinal absorption models. In single-dose pharmacokinetic studies of oral absorption, the primary concerns are with the extent of absorption and peak plasma or target tissue concentrations of the test substance. If the test vehicle affects gastric emptying, it may be necessary to use both fasted and non-fasted animals for pharmacokinetic studies. Blood (RBCs, plasma, and serum), urine, and feces are the most commonly collected samples. In addition, a few representative organ and tissue samples should be taken, such as liver, kidney, fat and suspected target organs. Sampling times should depend on the substance being tested and the route of administration. In general, an equal number of blood samples should be taken in each phase of the concentrationversus-time curve. Intravenous (IV) studies usually require much shorter, and more frequent, sampling than is required for oral dosing. Time spacing of samples will depend on the rates of uptake and elimination. In a typical IV study, blood and tissue samples are taken in a “powers of 2” series, i.e., samples at 2, 4, 8, 16, and 30 (32) minutes, 1, 2, 4, 8, and 16 hours. Similar coverage could be obtained with only seven time points by using a “powers of 3” series: 3, 9, and 30 (27) minutes, 1, 3, 9, and 24 (27) hours. Oral dosing studies usually extend to at least 72 hr. Such a sampling scheme would provide data coverage for evaluation of absorption, elimination, enterohepatic recirculation and excretion processes.

Toxicokinetics in Biomaterial and Device Safety Evaluation

269

The number of animals used in metabolism and pharmacokinetic studies should be large enough to reliably estimate population variability. In the case of rats and mice, tissue and/or blood sample size is usually the limiting factor; analysis of the substance may require 1 mL or more blood, but it is difficult to obtain multiple blood samples of this size from one animal. As a consequence, a larger number of animals is required (3–4 per time point, 7–9 time points) when small rodents are used. Such an approach has the advantage of allowing limited sampling of critical tissues (e.g., liver, fat) at each time point, an option which is usually unavailable with large animals. The use of humans and large animals generally permits collection of multiple (serial) blood samples. For outcrossing populations like humans and large animals, individual differences in the rates of biotransformation are likely to be greater than those of inbred rodent populations; under these circumstances, more samples/sex/ group may be needed to reliably estimate variability. Individual metabolism cages are recommended for collecting urine and feces in oral dosing studies. Excreta should be collected for at least five elimination half-lives of the test substance. When urine concentrations will be used to determine elimination rates, sampling times should be less than one elimination half-life (taken directly from the bladder in IV studies); otherwise, samples should be taken at equal time intervals. The results of the preliminary biotransformation/ kinetic study, together with the current regulatory metabolism studies and the 28- and 90-day studies should allow the selection of a relatively small number of appropriate tissues and/or fluids for monitoring purposes. Satellite groups of animals will provide the material for analysis. Methods must be developed to analyze non-radioactive test chemicals. Obviously it is important to monitor blood. It is accessible, convenient and, in certain circumstances, sequential sampling from the same animal may be important. The most useful aspect of blood is that the results can be compared with those obtained in man (see below). It is important, however, not to be constrained by this aspect. The most relevant tissues and body fluids should also be analyzed. These are target organs (if known) and indicator organs, tissues or fluids, i.e., those in which the concentration of pesticide or metabolite is a measure of that in the whole animal. In cases where distribution varies with dose (if shown in the preliminary study), a larger number of organs/tissues would be chosen for monitoring. Whether the parent drug or metabolite (or both) is chosen for analysis depends on the preliminary study. In principle, analysis for the parent compound should always be carried out; however, there are situations (e.g. rapid metabolism) when this is quite futile and a major retained metabolite should be used. Covalently bound metabolites are addressed below. Four occasions may be adequate for monitoring: (i) One month (equilibrium between intake of chemical and elimination of metabolites should be established; the time relates to the 28-day preliminary study). (ii) Three months (confirmation of results at one month; relates to the 90-day study). (iii) One year (coincides with the interim kill). (iv) Two years (effects of age; coincides with termination of study). Consideration should be given to the analysis of moribund animals.

270

Safety Evaluation of Medical Devices

In Vitro Studies In vitro measurements employing enzymes, subcellular organelles, isolated cells and perfused organs may be used to augment the dose-response information available from less extensive metabolic and pharmacokinetic studies. Because in vitro systems generally are less complex than whole animals, elucidation of a test compound’s metabolic pathways and the pathways’ kinetic characteristics may be facilitated. Such systems can be used to measure binding, adduct and conjugate formation, transport across cell membranes, enzyme activity, enzyme substrate specificity, and other singular objectives. Biochemical measurements that can be made using in vitro systems include: intrinsic clearances of enzymes in an organ or tissue, kinetic constants for an enzyme, binding constant, and the affinity of the test compound and its metabolites for the target macromoecules. The activity of a hepatic drug-metabolizing enzyme in vivo may be approximated by kinetic constants that are calculated from in vitro studies; when a first-order approximation is used, the ration of Vmax to Km is equal to the intrinsic clearance of the drug. In vitro measurements made using readily accessible tissues and body fluids from animals and man may also be useful in elucidating mechanisms of toxicity.

Analysis of Data Data from all metabolism and pharmacokinetic studies should be analyzed with the same pharmacokinetic model and results should be expressed in the same units. Concentration units are acceptable if the organ or sample size is reported, but percent of dose/organ is usually a more meaningful unit. In general, all samples should be analyzed for metabolites that cumulatively represent more than 1% of the dose. A variety of rate constants and other parameters can be obtained from IV and oral dosing data sets, provided that good coverage of the distribution, elimination, and absorption (oral dose) phase is available. Typical parameters calculated to characterize the disposition of a test substance are: half-lives of elimination and absorption; area under the concentration-versus-time curve (AUC) for blood; total body, renal and metabolic clearances (CI); volume of distribution (Vd); bioavailability (F); and mean residence and absorption times (MAT, MRT). Some of these parameters, such as half-lives and elimination rates, are easily computed from one another; the half-life is more easily visualized than the rate constant. Computation of oral absorption (ka) and elimination (E) rates is often complicated by the “flip-flop” of the absorption and elimination phases when they differ by less than a factor of 3. Because of these analysis problems, computation of absorption and elimination rates should not be attempted on the basis of oral dosing results alone. Blood-tissue uptake rates (kjl) can often be approximated from data at early (t < 10 minutes) time points in IV studies, provided that the blood has been washed from the organ (e.g., liver) or the contribution from blood to the tissue residue is subtracted (fat). High accuracy is not usually required since these parameters can be optimized to fit the data when they are used in more complex models. Tissueblood recycling rates (klj) and residence times can be computed from partition coefficients if estimates of uptake rates are available. Tissue/blood partition coefficients (Rjl) should be determined when steady-state has been achieved. Estimates based on samples obtained during the elimination phase following a single dose of the test substance may lead to

underestimates of this ratio in both eliminating and noneliminating tissues unless its half-life is very long. Correction of these values for elimination has been described by several authors. It may be important to determine the degree of plasma protein and red blood cell binding of the test substance; calculation of blood clearance rates using plasma or serum concentrations of the substance that have not been adjusted for the degree of binding may under- or over-estimate the true rate of clearance of the test substance from the blood. This is usually done through experiments in vitro. Two classical methods used in the analysis of pharmacokinetic data are the fitting of sums of exponential functions (two- and three-compartment mammillary models) to plasma and/or tissue data, and less frequently, the fitting of arbitrary polynomial functions to the data (non-compartmental analysis). Non-compartmental analysis is limited in that it is not descriptive or predictive; concentrations must be interpolated from data. The appeal of non-compartmental analysis is that the shape of the blood concentration-versus-time curve is not assumed to be represented by an exponential function and, therefore, estimates of metabolic and pharmacokinetic parameters are not biased by this assumption. In order to minimize errors in parameter estimates that are introduced by interpolation, a large number of data points that adequately define the concentration-versus-time curve are needed. Analysis of data using simple mammillary, compartmental models allows the estimation of all of the basic parameters mentioned above, if data for individual tissues are analyzed with one or two compartment models, and combined with results from two to three compartment analyses of blood data. "Curve stripping" analysis can be applied to such simple models through the use of common spreadsheet programs (i.e., LOTUS 1-2-3), as long as a linear regression function is provided in the program. Optimization of the coefficients and exponents estimated may require the use of more sophisticated software; a number of scientific data analysis packages such as RS/1 and SigmaPlot have the necessary capabilities. Specialized programs such as NONLIN, CONSAM, or SIMUSOLV will be needed when more complex models must be analyzed. Coefficients and exponents from mammillary models can be used to calculate other parameters; however, they should not be taken too literally, since mamillary models assume that all inputs are to a central pool (blood), which communicates without limitation into other compartments. This approach does not include details such as blood flow limitations, anatomical volumes or other physiological limits in the animal. Physiologically based pharmacokinetic models (PBPK) were developed to overcome the limitations of simple mammillary models. Physiologically based models describe the disposition of test substances via compartmental models which incorporate anatomical, biochemical, and physiological features of specific tissues in the whole animal. The types of information added include organ-specific blood flows, volumes, growth models, and metabolism rats. Metabolic parameters often are obtained from in vitro studies (i.e., enzyme reaction rates in cultured hepatocytes, plasma protein binding, etc.), while other parameters are becoming available as standard parameters in the literature. Parameters from mammillary models can be used to compute the value of parameters used in physiological pharmacokinetic

Chapter 19:

models, using tissue-specific blood flows, anatomical volumes, and other information (literature values). Estimation of parameters for a simple mammillary model is often the first data reduction step in creating a physiological model. Because PBPK models are based on physiological and anatomical measurements and all mammals are inherently similar, they provide a rational basis for relating data obtained from animals to humans. Estimates of predicted disposition patterns for test substances in humans may be obtained by adjusting biochemical parameters in models validated for animals; adjustments are based on experimental results of animal and human in vitro tests and by substituting appropriate human tissue sizes and blood flows. Development of these models requires special software capable of simultaneously solving multiple (often very complex) differential equations, some of which were mentioned above. Several detailed descriptions of data analysis have been reported.

Use of Data from Metabolism and Pharmacokinetic Studies Information from metabolism and pharmacokinetic studies can be used in the design and analysis of data from other toxicity studies. Some examples are described below. Design of Toxicity Studies: The concentration-versus-time curve, peak and steady-state concentrations of the test substance in blood or plasma provide information on the distribution and persistence of the substance in the animal which may suggest essential elements in the design of the toxicity studies. For example, when metabolic and pharmacokinetic studies indicate that the test compound accumulates in the bone marrow, long-term toxicity tests should include evaluation of the test compound’s effect on hematopoietic function and morphology. If a test compound is found to accumulate in milk, an investigator may need to plan to perform reproductive toxicity studies with in utero exposure and a nursing phase (cross-fostering study). In addition, information from metabolic and pharmacokinetic studies can be used to predict the amount of test compound that enters biological compartments (tissues, organs, etc.) that may not suffer a toxic insult but may serve as depots for indirect or secondary exposure.

Whole Body Autoradiography Autoradiography is the production of an image in a photographic emulsion by the emission from a radioactive element. The term autoradiography is preferred to that of radioautography. Prefixes are added to words to further classify the concept. Therefore, the process is “auto-” radiography for a “self-” radiograph and not a “radio-” autograph or one’s transmitted signature (Waddell, 1972). Whole Body Autoradiography (WBA) has been used with increasing frequency as a means of identifying tissues which concentrate test substances. This technique allows a small number of animals (5–10) to be used for screening purposes with a minimal investment in manual labor. FDA encourages the use of WBA with IV dosing as a means of screening and selecting tissues of greatest relevance for later oral dosing studies. Animals used for WBA should generally be sacrificed during primary consideration in selecting specific tissues. The most comprehensive technique currently available for the initial survey of the distribution of a drug is that of whole body autoradiography. The species of animals used include mice, rats, hamsters, monkeys, pigs, dogs, and ferrets.

Toxicokinetics in Biomaterial and Device Safety Evaluation

271

The most widely used animal has been the mouse, which has the advantages of requiring less isotope and being easier to section. The animals are anesthetized and then frozen by immersion at various times after administration of the labeled compound in hexane or acetone cooled with dry ice. Since the freezing in the interior of the animals occurs slowly, large ice crystals form within these tissues, hence subcellular localization of compounds is not possible. The selection of times for freezing an animal after injection of a drug must be based on the information available on the rate of elimination of the compund from the animal by metabolism and excretion. In general, a geometric increase in time intervals is most useful. In order to have time intervals for comparison, we routinely have employed freezing times which are approximately multiples of three, namely 2 minutes, 6.5 minutes, 20 minutes, 1 hour, 3 hours, 9 hours, and 24 hours. In certain cases, rapid elimination of the drug by the kidneys must be circumvented by ligation of the renal pedicles to avoid apparent localization from failure of the agent to reach equilibrium. An example of this is given in Section IV, A for urea-14C in pregnant mice. The animal is frozen into a block of carboxymethylcellulose ice on the microtome stage. Although the Jung, type K, microtome has been used, the Leitz, model 1300, sledge microtome is more suitable, for its smaller size allows it to be mounted in an ordinary commercial freezer instead of a walk-in freeze room. The microtome stage must be designed for mounting in the vice by the front end of the stage. Sections from 5 µm to approximately 80 µm thick are taken onto #800 Scotch tape (Minnesota Mining & Mfg. Co.). Before removal from the freezer, the sections must be allowed to dry thoroughly so that no ice remains which can melt and allow movement of the isotope. After drying, if covered to prevent condensation of moisture on the sections, the sections may be transferred from the freezer to room temperature. X-ray films which produce the most satisfactory autoradiograms are Kodak industrial type AA and Gevaert Structurix D-7. Both are fine grain films which have been demonstrated not to produce chemical artifacts. Approximately six times faster, Kodak No Screen and Kodirex may be used for rapid screening and timing of autoradiograms. However, the occasionally produce artifacts and should not be relied on for interpretation. Some investigators have used photographic emulsions such as Ilford G-5, 10 µm thick, pre-applied to glass plates. The increased cost and likelihood of breakage, however, hardly justifies the small improvement in resolution for whole-body sections. Exposure of the photographic emulsion by the radioactivity of the tissue section should be at freezer temperatures to prevent autolysis of the tissue. After exposure of the x-ray film, sections with isotopes which have a long halflife may be placed against fresh x-ray film for additional sets of autoradiograms with either a longer or shorter exposure time. This procedure is useful for revealing relative concentrations of radioactivity for areas that have either very high or very low concentrations after the first exposure. When no further autoradiograms are needed, the section can be stained with histological dyes to verify localizations of radioactivity. Compounds that fluoresce under ultraviolet light can be visualized in the tissue sections and their locations recorded with color film. Whole-body tissue sections can be

272

Safety Evaluation of Medical Devices

used for histochemical localizations for comparison with the autoradiograms. Furthermore, the areas can be removed, extracted, and the extract chromotographed to identify the chemical nature of the radioactivity revealed by the autoradiogram. Although the whole-body technique will allow localization of an increased concentration of an isotope in a tissue or occasionally a cell type, other techniques must be used for single cells and subcellular localization. A nuclear tract plate is prepared by dipping the plate in a 12% solution of glycerine in absolute ethyl alcohol and allowing it to drain for 10 minutes in a vertical position before approximating the section on tape. After the emulsion is exposed, soaking in xylene removes the tape but leaves the section attached to the nuclear tract plate. The Ilford G-5 nuclear tract plates with 10 µm emulsions are most satisfactory. The increased resolution gained by the finer grained Ilford K and L emulsions is warranted only for tissues that are well preserved and relatively free of ice crystal artifacts. Kodak NTB emulsions seem to produce more pressure artifacts than the Ilford plates. Comparison of various techniques of autoradiography for diffusible compounds clearly demonstrates that no solutions can be used in processing the tissue. These investigators have dried thin sections of liver and uterus at temperatures below –60∞C. These freeze-dried sections were dry mounted on microscope slides which had been precoated with either Kodak NTB-3 or NTB-10 emulsion. Other techniques which thawed the frozen section, embedded the tissue in paraffin or dipped the section in liquid emulsion, were demonstrated to translocate diffusible compounds. Many other similar attempts have been and are currently being made to localize diffusible compounds by autoradiography at the electron microscope level.

PHYSIOLOGICALLY-BASED PHARMACOKINETIC (PBK) MODELING Pharmacokinetic modeling is the process of developing mathematical explanations of absorption, distribution, metabolism, and excretion of chemicals in organisms. Two commonly used types of compartmental pharmacokinetic models are (a) data-based and (b) physiologically based. The data-based pharmacokinetic models correspond to mathematical descriptions of the temporal change in the blood/tissue level of a xenobiotic in the animal species of interest. This procedure considers the organism as a single homogeneous compartment or as a multicompartmental system, with elimination occurring in specific compartments of the model. The number, behavior, and volume of these hypothetical compartments are estimated by the type of equation chosen to describe the data, and not necessarily by the physiological characteristics of the model species in which the blood/tissue concentration data were acquired. Whereas these data-based pharmacokinetic models can be used for interpolation, they should not be used for extrapolation outside the range of doses, dose routes, and species used in the study on which they were based. In order to use the data-based models to describe the pharmacokinetic behavior of a chemical administered at various doses by different routes, extensive animal experimentation would be required to generate similar blood-time course data under respective conditions. Even within the same species of animal, the time-dependent nature of critical biological

determinants of the disposition (e.g., tissue glutathione depletion and resynthesis) cannot easily be included or evaluated with the data-based pharmacokinetic modeling approach. Further, due to the lack of actual anatomical, physiological, and biochemical realism, these data-based compartmental models cannot easily be used in interspecies extrapolation, particularly to predict pharmacokinetic behavior of chemicals in humans. These various extrapolations, which are essential for the conduct of dose–response assessment of chemicals, can be performed more confidently with a physiologically based pharmacokinetic modeling approach. This chapter presents the principles and methods of physiologically based pharmacokinetic modeling as applied to the study of toxicologically important chemicals. PBPK modeling is the development of mathematical descriptions of the uptake and disposition of chemicals based on quantitative interrelationships among the critical biological determinants of these processes. These determinants include partition coefficients, rates of biochemical reactions, and physiological characteristics of the animal species. The biological and mechanistic basis of the PBPK models enables them to be used, with limited animal experimentation, for extrapolation of the kinetic behavior of chemicals from high dose to low dose, from one exposure route to another, and from test animal species to people. The development of PBPK models is performed in four interconnected steps: model representation, model parameterization, model stimulation, and model validation. Model representation involves the development of conceptual, functional, and computational descriptions of the relevant compartments of the animal as well as the exposure and metabolic pathways of the chemical. Model parameterization involves obtaining independent measures of the mechanistic determinants, such as physiological, physicochemical, and biochemical parameters, which are included in one or more of the PBPK model equations. Model simulation involves the prediction of the uptake and disposition of a chemical for defined exposure scenarios, using a numerical integration algorithm, simulation software, and a computer. Finally, the model validation step involves the comparison of the a priori predictions of the PBPK model with experimental data to refute, validate, or refine the model description, and the characterization of the sensitivity of tissue dose to changes in model parameter values. PBPK models after appropriate testing and validation can be used to conduct extrapolations of the pharmacokinetic behavior of chemicals from one exposure route/scenario to another, from high dose to low dose, and from one species to another. The PBPK model development for a chemical is preceded by the definition of the problem, which in toxicology may often be related to the apparent complex nature of toxicity. Examples of such apparent complex toxic responses include nonlinearity in dose–response, sex/species differences in tissue response, differential response of tissues to chemical exposure, qualitatively and/or quantitatively difference responses for the same cumulative dose administered by different routes/scenarios, etc. In these instances, PBPK modeling studies can be utilized to evaluate the pharmacokinetic basis of the apparent complex nature of toxicity induced by the chemical. One of the values of PBPK modeling, in fact, is that accurate description of target tissue dose often resolves behavior that appears complex at the administered dose level. The principal application of PBPK models is in the prediction of the target tissue dose of the toxic parent chemical

Chapter 19:

or its reactive metabolite. Use of the target tissue dose of the toxic moiety of a chemical in risk assessment calculations provides a better basis of relating to the observed toxic effects than the external or exposure concentration of the parent chemical. Because PBPK models facilitate the prediction of target tissue dose for various exposure scenarios, routes, doses, and species, they can help reduce the uncertainty associated with the conventional extrapolation approaches. Direct application of modeling includes: • • • •

high-dose/low-dose extrapolation route–route extrapolation exposure scenario extrapolation interspecies extrapolation.

BIOLOGICALLY DERIVED MATERIALS The progress and products of biotechnology have brought some new challenges to the assessment of pharmacokinetics and toxicokinetics. While the reasons for needing this data (demonstrating, exposure, displaying dose dependency, correlating any findings of toxicity to exposure and determining steady state for systemic agent levels) are certainly as compelling as with traditional drugs, there are a whole set of special problems involved. These special concerns for biologically derived products are:

Toxicokinetics in Biomaterial and Device Safety Evaluation

273

The upshot of these points is that it may not be practical to follow established guidelines for ADME evaluation. Binding proteins, immunoreactive metabolites and antibodies could interfere with the immunoassays used to measure the activity of biotechnologically-derived pharmaceuticals. The link between immunoreactivity and pharmacological activity may be difficult to establish, making the data difficult to interpret. In radiolabeled distribution studies, if the label alters the physicochemical and biological properties of the test material, its pharmacokinetic behavior may change. These analytical difficulties may preclude accurate characterization of the distribution, metabolism and excretion of a protein. AUC and Cmax are commonly measured to identify safety ratios for new chemical entities. Since the analytical methods used for biotechnologically-derived pharmaceuticals may lack specificity, a clinical marker of biological activity or efficacy may sometimes be more appropriate than exposure data. It is therefore essential that before pivotal (repeat dose) preclinical studies are initiated, bioanalytical assay development must be completed. This has to cover potential test species, normal (and diseased humans). The assays must be validated in the sampling matrix of the toxicity test species, and one should also develop suitable assays for antibodies to the test article.

Assay Sensitivity/Specificity • Needs to be at 1 ng/ml or lower • Cross reactivity to native protein may confound results • If test article is the same as native protein, how do you tell the difference? • Western blot can be used to demonstrate specificity • Antibody interference may occur with assay. Low Systemic Levels • Rapid metabolism — Metabolites may be endogenous proteins or amino acids • Extensive metabolism — Metabolites may be incorporated into cell structures rapidly • Rapid distribution • Rapid hepatic clearance • Route of administration may bypass systemic circulation — SC — ICV; ITs Endogenous Protein • May cross react and lead to false positive blood levels • Can radiolabel to tell the difference between administered molecule and endogenous molecule — However, the label may lead to different distibution — What is the specific activity if diluted with unlabeled endogenous material? Sample Volume • May need to be large to increase sensitivity but also • May need to be small because of competing assays — Immune factors [antibodies, globulins] — Hormones — Disease state modifiers — In humans, concomitant medications • And available test material supply will be very limited in early development.

REFERENCES AAMI (2003). ISO 10993-16: Toxicokinetic study design for degradation products and leachables for medical devices. AAMI, Arlington, VA. Andersen, M.E., Clewell, H.J. III, Gargas, M.L., Smith, F.A. and Reitz, R.H. (1987). Physiologically-based pharmacokinetics and the risk assessment process for methylene chloride. Toxicol. Appl. Pharmacol., 54: 100–116. Bronaugh, R.L. (1998) Methods for in vitro percutaneous absorption. In F.N. Marzulli and H.I. Maibach (Eds.), Dermatotoxicology Methods. Taylor & Francis, Philadelphia. Caldwell, W.S., Byrd, C.D., DeBethizz, J.D. and Brooks, P.A. (1994). Modern instrumental methods for studying mechanisms of toxicity. In: A.W. Hayes (Ed.), Principles and Methods of Toxicology, Raven Press, New York. Connally, R. and Anderson, M. (1991). Biologically based pharmacokinetic models: Tools for toxicological research risk assessment. Annu. Rev. Pharmacol. Toxicol., 31: 503–523. (CYP3A) activity in human liver microsomes. Drug Metab. Dispos., 26: 299–304. FDA (2000). Redbook II, pp. 138–148. Gabrielsson, J. and Weiner, D. (1997). Pharmacokinetic/Pharmacodynamic Data Analysis: Concepts and Applications. Apotekarsociateten, Stockholm. Gad, S.C. and Chengelis, C.P. (1997) Acute Toxicology Testing, 2nd Edn. Academic Press, San Diego, CA. Gad, S.C. and Chengelis, C.P. (Eds.) (1992). Animal Models in Toxicology. Marcel Dekker, Inc., New York. Garner, C. and Matthews, H. (1998). The effect of chlorine substitution on the dermal absorption of polychlorinated biphenyls. Toxicol. Appl. Pharmacol., 149: 150–158. Goldstein, A., Aronow, L. and Kalman, S. (1974). Principles of Drug Action: The Basis of Pharmacology. John Wiley & Sons, New York. Griffin, R., Godfrey, V., Kim, Y., and Burka, L. (1997) Sex-dependent differences in the disposition of 2,4-dichlorophenoxyacetic acid in Sprague-Dawley rats, B6C3F1 mice and Syrian hamsters. Drug Metab. Dispos., 25: 1065–1071.

274

Safety Evaluation of Medical Devices

Holland, J., Kao, M. and Whitaker, M.J. (1984). A multisample apparatus for kinetic evaluation of skin penetration in vitro: The influence and metabolic status of the skin. Toxicol. Appl. Pharmacol., 72: 272–280. International Programme on Chemical Safety (IPCS) (1986). Principles of toxicokinetic studies. Environmental Health Criteria, 57, World Health Organization, Geneva. Jollow, D.J., Roberts, S., Price, V., Longacre, S. and Smith, C. (1982). Pharmacokinetic considerations in toxicity testing. Drug Metab Rev., 13: 983–1007. Keberle, G., Brindle, S.D. and Greengard, P. (1971). The route of absorption of intraperitoneally administered compounds. J. Pharmacol. Exp. Ther., 178: 562–566. La Du, B., Mandel, H. and Way, E. (1972). Fundamentals of Drug Metabolism and Drug Disoposition. The Williams & Wilkins Co., Baltimore, MD. Lin, C.S., Shoaf, S.E. and Griffiths, J.C. (1992). Pharmacokinetic data in the evaluation of the safety of food and colour additives. Reg. Toxicol. Pharmacol., 15: 62–72. Lu, H., Wang, J., Chan, K. and Young, D. (1998). Effects of phenobarbital of stereoselective metabolism of ifosfamide in rats. Drug Metab. Disp. 26: 476–482. Meyer, U.A. (1994). The molecular basis of genetic polymorphisms of drug metabolism, J. Pharm. Pharmacol. (suppl 1): 409–415. Monostory, K., Jemnitz, K., Vereczkey, L. and Czira, G. (1997). Species differences in metabolism of panomifene, an analogue of tamoxifen. Drug Metab. Dispos., 25: 1370–1378. Monro, A.M, (1990). Interspecies comparisons in toxicology: The utility and futility of plasma concentrations of the test substance. Reg. Toxicol. Pharmacol., 12: 137–160. Mugfor, C. and Kidderis, G. (1998). Sex-dependent metabolism of xenobiotics. Drug Metab. Reviews, 441–498. O’Reilly, W.J. (1972). Pharmacokinetics in drug metabolism and toxicology. Can. J. Pharm. Sci., 7: 66–77. Organization for Economic Cooperation and Development (OECD). Guidelines for Testing of Chemicals—No. 417. Toxicokinetics, OECD Publications, Paris. Pratt, W.B. and Taylor, P. (1990). Principles of Drug Action: the Basis of Pharmacology, 3rd Edn. John Wiley & Sons, New York.

Renwick, A.G. (2000). Toxicokinetics, In: B. Ballantyne, T. Marrs and T. Syversen (Eds.), General and Applied Toxicology. Grove’s Dictionaries, Inc., New York. Smith, D. (1991) Species differences in metabolism and pharmacokinetics: are we close to an understanding? Drug Metab. Rev., 23: 355–373. Shapiro, J. (1981). Radiation Protection, 2nd Edn. Harvard University Press, Cambridge, MA, pp. 12–18. Shargel, L. and Yu, A.B.C. (1999). Applied Biopharmaceutics and Pharmacokinetics. Appleton & Lange, New York. Smith, D.A., Humphrey, M.J. and Charuel, C. (1990). Design of toxicokinetic studies. Xenobiotica, 20: 1187–1199. Travis, C.B. (1998). Pharmacokinetics. In: Traves, C.B. (Ed.), Carcinogen Risk Analysis. Contemporary Issues in Risk Analysis, vol. 3. Plenum Press, New York. Tse, F.L.S. and Jaffe, J.M. (1991). Preclinical Drug Disposition. Marcel Dekker, Inc., New York. Vinegar, A. and Jepson, G. (1996). Cardiac sensitization thresholds of halon replacement chemicals in humans by physiologically based pharmacokinetic modeling. Risk Analysis, 16: 571–579. Waddell, W.J. (1972). Autoradiography in drug dispositions studies. In: B.N. La Du, H.G. Mandel and E.L. Way, (Eds.), Fundamentals of Drug Metabolism and Drug Disposition. Williams and Wilkins, New York, pp. 505–514. Wang, Y.M. and Reuning, R. (1994). A comparison of two surgical techniques for the preparation of rats with chronic bile duct canulae for the investigation of enterohepatic circulation. Lab. Animal Sci., 44: 479–485. Wartak (1983). A modern approach to individualized drug therapy. Clinical Pharmacology and Therapeutics Series, Vol. 2. Praeger Publishers CBS Educational and Professional Publishing. Weissinger, J. (1989). Nonclinical pharmacologic and toxicologic considerations for evaluating biologic products. Reg. Toxicol. Pharmacol., 10: 255–263. Welling, P.G. and De La Iglesia, F.A. (1993). Drug Toxicokinetics. Marcel Dekker, Inc., New York. Yacobi, A., Skelly, J.P. and Batera, V.K. (1989). Toxicokinetics and New Drug Development. Pergamon Press, New York.

20 Case Histories And Problem Resolution

Public and societal perceptions of the safety of medical devices (and therefore, to a large extent, regulatory requirements to establish safety prior to marketing) have been shaped by the “known” history of the issue. Unfortunately, the publicly known history is primarily based on what has appeared in the popular press, which continues to mean primarily a litany of problems. The last thirty years have seen six cases of significance in the public mind. By examining these six cases, their causes, the regulatory responses to them, and solutions to the problems they revealed, the medical device researcher and developer can learn some valuable lessons. After reviewing these cases, biocompatability problem solving approaches and the means to place actual risks in context, risk assessment will be considered.

Both DEHP and MEHP at relatively low concentrations cause liver damage (Lake et al., 1976; Nikonorow et al., 1973), and both compounds accumulate in the fatty tissue of the liver. At high levels in the diet, DEHP causes liver cancer in rodents (Sheftel, 1994). DEHP was subsequently discontinued as a plasticizer in blood bags after a great deal of concern and study. Other additives (benzothiazoles and other vulcanization accelerators in rubbers, for example) and contaminants (such as nickel and chromium) have subsequently also been found to migrate into solutions in contact with them (Lazarus, 1980; Petersen et al., 1981; Meek and Pettit, 1985; Roster and Gruenwald, 1986). It should be expected that there will be migration of some amount of any residual monomer, additive, or contaminant in a plastic or elastomer used in devices into any fluid (or mucous membrane tissue) with which there is contact. If any of these leachable materials is also toxic, there may be a risk to patients—real or perceived. It is essential to know the composition of materials used in devices at least qualitatively and should one of the potentially leachable materials present a hazard, it becomes essential to know how much is present and (if the amount present is sufficient) how much is leached into solutions and the amount of time that the device will be in contact. The issue has been given new life yet again in 2006/2007 due to our technological ability to measure increasingly small amounts of phthalates in human tissue and finer degrees of change in biological tissues and processes for DEHP (and in 2008, bisphenol A) endocrine disruption effects potentially leading to effects on reproduction and development of sexual organs and function in children. As new research has suggested that lower levels of DEHP can potentially lead to reproductive effects, the level of conern as to their use in (and leaching from) medical devices has increased again. The reader is referred to the draft ISO 29741 (2007) guidance for an excellent review of the available human and animal data. This draft guidance proposes to set, based on the risk assessment method promulgated in ISO 10993-17, a maximum TIs (tolerable daily patient exposure levels for DEHP leaching from medical devices) of 0.6 (parenteral exposure), 0.06 ((oral exposure), and 0.003 (inhalation exposure) mg/kg/day for DEHP. Also based on the same animal and human studies, there has been increased concern on the part of devices companies as to potential fines, awards and actions in California under Proposition 65. To determine if marketed devices meet these pending ISO TIs (and therefore provide a basis for not being subject to actions under Proposition 65), a three tier approach is recommended.

PHTHALATE AND OTHER RESIDUALS LEACHING FROM DEVICES Since the Second World War, plastics have come to be used as the major material in the fabrication of disposable medical devices. Starting in the early 1960s, reports began to appear about the toxicity of various chemical additives used in plastic formulations (Guess and O’Leary, 1969; Autian, 1972; Eckardt and Hinden, 1973), particularly those in polyvinyl chloride, which were found to readily leach out into blood and IV solutions stored in or passing through them. Much of this attention came to focus on the phthalate plasticizers, particularly di-2-ethylbenzyl phthalate (DEHP). The plasticizers largely serve to increase the flexibility of PVC products, such flexibility being essential for certain (large volume) types of products such as tubing and blood or fluid bags. Studies clearly established that there was a time dependent leaching of DEHP from polyvinylchloride (PVC) bags into blood and blood components stored in such bags, even at refrigerator temperatures (4∞C) used for prolonged storage of the blood materials (Peck et al., 1979; Sasakawa and Mitomi, 1978; Jaeger and Rubin, 1973; Mayer et al., 1972; Miripal, 1977). Such migration also occurred from other PVC devices into blood, intravenous fluids, and dialysis fluids which also had significant contact time with such plastic fluids. The two populations with the greatest resulting exposures were hemophiliacs (who received frequent transfusions of blood) and kidney dialysis patients (who received treatment three or four times a week). Studies in monkeys (Peck et al., 1979), rats (Daniel and Bratt, 1973; Carter et al., 1974), and stored human blood (Rock et al., 1978) also established that significant portions of the DEHP was hydrolyzed to mono-2-ethyl-hexylphthalate (MEHP). The amount of lipids in the blood strongly influenced the rate of leaching.

(1) Determine by calculation whether the level of DEHP used in device polymer manufacture has the potential, if all extracted from the device, to exceed the appropriate 275

276

Safety Evaluation of Medical Devices

limit for the device. This is a paper exercise and will give the highest numbers, but if this worst case does not present the potential to have levels of DEHP extracted that would exceed the limit, no real world potential exists and the device may be considered “qualified.” (2) Perform extractions in saline and a 10% ethanol solution at 37°C for 72 hours at ISO compliant mass/volume ratios. Do not cut up the device(s) in question before extraction. Analysis of the solutions and calculate total DEHP extracted. Is the total below the allowance level? If so, you are done. If not ... (3) Do implantation study in appropriate animal models, collecting blood samples and analyzing to determine how much DEHP is relased. Is this total below the TI level?

DALKON SHIELD Efforts to prevent pregnancy by placing a foreign object in a woman’s uterus date back many centuries. Devices made of glass, ivory, wood, wool, silkworm gut, silver, gold, copper, zinc, and pewter in a variety of sizes and shapes have all been used for this purpose. The reason for the contraceptive effect of an intrauterine device (IUD) that does not create a barrier to the entry of sperm into the uterus has never been satisfactorily established. One theory is that in reaction to the device the body produces white cells that destroy the sperm or the fertilized egg. Another is that the IUD somehow interferes with the ability of the fertilized egg to attach itself or to remain attached to the wall of the uterus. Infection and injury have historically been associated with IUDs, and they were not widely used until the late 1960s. Then, growing dissatisfaction with the adverse effects of birth control pills, together with mounting concern over the “population explosion,” caused renewed interest in the development of a safe IUD as a means of contraception. The possibility was enhanced by two scientific advances. One was the development of antibiotics that provided a method of treating those uterine infections that did occur. The other was the invention of malleable, inert plastic, which could be compressed during insertion and then regain its intended shape, and which would not break down or chemically interact with bodily fluids. The Dalkon Shield intrauterine device was designed in 1968 by Hugh Davis, a gynecologist on the faculty of the Johns Hopkins Medical School, and Irwin Lerner, an electrical engineer and part-time inventor. In 1964 John Hopkins had opened a family planning clinic under Davis’s supervision and through this clinic Davis was able to test various newly developed IUDs, mostly on low-income, black women from Baltimore. A major design problem with most of these IUDs was the frequency with which they were expelled as a result of the body’s natural efforts to rid itself of a foreign object. The Dalkon Shield was specifically designed to control the problem of expulsion. It is a piece of flexible plastic, about three-quarters of an inch across, shaped like a shield to approximate the shape of the uterine cavity. The device has four or five prongs jutting out of each side at a downward angle. Because of the direction of the prongs, the device resisted expulsion. However, the same prongs often made insertion or removal painful and difficult, and they were responsible for the tendency of the device to embed itself in the uterine wall or to perforate the uterus. A more serious design problem of the Dalkon Shield related to its string. One end of the string was tied to the bottom

of the plastic shield and the other end was tied in a knot. When the device was in the uterus, the string passed through the cervix into the vagina. Its purpose was to allow the woman to check that the device was in place and to assist in its removal. “Tailstrings” designed to serve these purposes are found on all IUDs, but on IUDs other than the Dalkon Shield the tailstring is made of a single plastic filament—a “monofilament”—in order to avoid the absorption of moisture, and with it bacteria. The entry of bacteria from the vagina into the uterus is a major cause of dangerous and potentially life-threatening uterine infections, known as pelvic inflammatory disease, or PID. Frequently, a hysterectomy must be performed to overcome PID, and even when a hysterectomy is not necessary, infertility can result. Davis and Lerner did not use a monofilament string on their new Dalkon Shield. The monofilament strings they tested were either too weak and would break during removal because of the resistance offered by the prongs or too stiff and would cause discomfort to the man during intercourse. Instead, Davis and Lerner used a multifilament string, named Supramid, which was composed of hundreds of tiny nylon strands encased in a nylon sheath. Supramid was made by a company in West Germany and used primarily in the repair of horse tendons. Davis and Lerner hoped that the sheath would prevent vaginal fluid from entering the spaces between the strands and moving up into the uterus, a phenomenon called “wicking” (Sobol, 1991). For two reasons, the sheath did not do its job. First, inexplicably, Davis and Lerner failed to seal its ends. Fluid from the vagina could enter the open end at the bottom of the sheath and wick up past the knots to the open end at the top inside the uterus. Indeed, the nylon sheath made the problem of wicking worse because it shielded the bacteria inside the sheath from the antibacterial action of the mass of viscous fluid in the cervix, known as the cervical plug. Second, the sheath developed holes either in the initial tying process or as a result of decomposition after the IUD was in place. These holes allowed the bacteria to escape into the uterus without even reaching the top of the string (Tatum et al., 1975). The two men formed the Dalkon Corporation, which in February 1970 began commercial sales of the shield. Marketing rights were purchased by the A.H. Robins Company in June of 1970 (Sivin, 1993). Some 4.5 million of the devices were distributed throughout the world, mostly between 1971 and 1975 (Byrne, 1992). In 1973, the U.S. Food and Drug Administration (FDA) conducted an enquiry into the Dalkon Shield device. In 1974 the company ceased manufacturing in the United States but did not stop marketing elsewhere in the world until later dates. Stocks were still available outside the United States early in the 1980s. People started taking legal action against the manufacturer soon after the Dalkon Shield was marketed, but Robins declared bankruptcy in 1985 to shield themselves from the onslaught of litigation. A trust was set up, and with the publicity ordered by the bankruptcy court, claims flooded in from all around the world. Fifteen thousand American women sued Robins before the company declared bankruptcy in August of 1985, and a $3.3 billion settlement was reached in 1988 (Nocera, 1995). These suits were due to unacceptable rates of pregnancy, pelvic inflammatory disease, and death in women who used it. The manufacturer knew of these defects in March of 1972, but a least one million device insertions are believed to have occurred after that. Thousands of women have reportedly

Chapter 20:

suffered severe and debilitating medical problems, including pain and bleeding, uterine perforations, pelvic inflammatory disease, septic abortion, ectopic pregnancies and unplanned pregnancy. Women have become sterile, many having hysterectomies in their twenties, and many spending the rest of their lives on hormonal replacement therapy. By 1976, 17 women were known to have died in the United States following the use of the Shield. A side-effect of these events was a public belief that all IUDs were unsafe, and a business belief that all IUDs were litagens. As a result, this form of birth control largely disappeared from the marketplace (Mumford and Kessel, 1992), despite evidence that the Dalkon Shield may have been acceptable, safe and effective when properly inserted, and no evidence that the other two significant forms of IUDs (Lippes Loop and the Cooper IUDs) were not effective or safe (Sivin, 1993). Premarket clinical efficacy safety testing of the Dalkon Shield was insufficient, manufacturing quality control was poor, and once widely sold, clinical use was poorly controlled. Additionally, preclinical safety testing was almost nonexistent and did not reflect any consideration of device design, intended use, or probable misuse.

TALC ON GLOVES AND CONDOMS Particles of talc have been identified on both the surface of surgeons’ rubber gloves and on latex condoms, as well as in sections of tissue classified as “starch” or “foreign body” granulomas. The use of gloves to prevent contamination of operative wounds was accidental. The first person to use them in the United States was the wife of Dr. W. S. Halsted, who used gloves in the operating room to protect her hands and arms from the carbolic acid (in which the instruments were kept) which caused her a severe dermatitis. Joseph Bloodgood has the distinction of being the first surgeon to wear gloves in every operation starting in 1896. Halstead many years later commented on the usefulness of gloves and described how they were only a minimal hindrance (Singh et al., 1974). The earlier preparation for dry lubrication of surgical gloves was Lycopodium spores, and later, talcum powder. These two agents fell into disrepute when it was reported that a definite foreign body reaction to talc occurred in numerous regions of the body. In 1947, Lee and Lehman demonstrated experimentally the advantages of a cornstarch derivative and its apparent safety in human tissues. It was widely accepted, and in the 1950s a cornstarch derivative (Biosorb) rapidly replaced talc as a surgical glove powder. The adverse effects of cornstarch in patients were first recorded by Snierson and Woo (1955), who reported two cases with incisional granulomas. Since then, numerous reports of granuloma formation due to cornstarch have been recorded in various anatomic sites, such as the abdominal cavity, pulmonary parenchyma, pleura, paranasal sinuses, mastoid, testes, and biopsy scars. Granuloma formation has been reported following percutaneous of cornstarch material up the female genital tract following a pelvic examination. Numerous experimental studies using the abdominal cavity of mice, dogs, rabbits, and guinea pigs have been carried out to evaluate cornstarch as a glove powder lubricant. Injections into the soft tissues of the backs of hamsters and

Case Histories and Problem Resolution

277

mice have produced cornstarch granulomas. Single reports have also been mentioned in experiments carried out on the brain, spinal cord, and peripheral nerves by Wise, as well as on lung, knee joint and tendon (Postlethwaite et al., 1949). A single case report of synovial involvement has appeared in the literature, as a letter to the editor of the Journal of the American Medical Association. It involved a patient who developed a granuloma at the site of excision of a ganglion from the wrist. Postlethwaite implanted a “pinch” of glove powder in the knee joints of six dogs, and found no gross changes in the synovium at the end of 14 and 28 days, but did find microscopic tissue reaction. Only rarely were starch granules seen and there was no evidence of granuloma formation. The degree of intra-articular inflammation is directly related to the amount of powder per surface area of the joint. In an abdominal operation, the entire hand of the surgeon may be introduced into the body cavity. The recognition that the use of talc as a surgical glove powder was responsible for the postoperative development of peritoneal adhesions and tissue granulomas at the site of operation (Lichtman et al., 1946; Eismann et al., 1947) stimulated the search for an inert absorbable substitute, and resulted in its eventual replacement by a starch powder (Lee and Lehman, 1947; Postlethwaite et al., 1949), which initially appeared to produce no ill effects within the peritoneal cavity. There have, however, been numerous reports over the past two decades suggesting that the use of starch powder could also result in a granulomatous reaction similar to that produced by talc (Snierson and Woo, 1955; McAdams, 1956; Nash, 1971; Neelly and Davis, 1971; Cohen and Safaie-Shirazi, 1973). Starch granuloma peritonitis secondary to the glove powder reaction has now become recognized as a clinical entity (Sobel et al., 1971), and Taft et al. (1970) suggested that the severity of the condition may be related to the type of starch used. The starch powder has also been shown to produce granulomas and adhesions in experimental animals (Lee et al., 1952; Perper et al., 1971; Ignatius and Hartmann, 1972). However, recent reports indicate that starch particles may not always have been responsible for the development of the granuloma peritonitis and that contaminants of the dusting powder used in the surgeons’ gloves must be taken into consideration (Henderson et al., 1975). Similar reports have appeared associating dry lubricants on condoms with adverse tissue responses. Given the longstanding knowledge of the systemic effects of particulates introduced into the body cavities, it should have been clear that any solid, poorly soluble, or insoluble particle introduced into the body or a body cavity is likely to cause a foreign body response mediated reaction in the body. Similar responses are seen to the wear particles produced from artificial joint implants.

TOXIC SHOCK SYNDROME The tampon was first introduced in 1936. Since then its use in menstruating women has expanded greatly, with an estimated 73 million women using it up to five times a day, five days a month for up to 38 years (Houppert, 1995). Though there were reports in the literature as early as 1966 (Maguire, 1966; Orm and Beck, 1981) that tampons were associated with vaginal infections (“tampon vaginitis”), the cases were generally mild and transitory. There was no history of serous medical problems (requiring hospitalization)

278

Safety Evaluation of Medical Devices

or of deaths associated with tampon use until 1978, some three years after Proctor and Gamble first test-marketed the Rely superabsorbent tampon in 1974. FDA began to monitor negative reaction to tampons in 1974 through the Device Experience Network (DEN). Complaints were registered by consumers as well as health professionals; complaints which described experiences with mucosal alterations in the vaginal area were drying, layering, lacerations, micro-ulcers, hemorrhaging, and dermatitis. In several instances doctors stated that they had reported these experiences to the manufacturers and had been ignored. In 1977 four cases of tampon-induced vaginitis were described by Barrett, who wrote that these should “alert the clinician to consider the possibility of tampon-induced pressure or chemical irritation in the differential diagnosis of a vaginal ulceration . . . second, the patients could have manifested an idiosyncratic reaction to the chemicals in the deodorized tampon.” In November 1978, toxic shock syndrome (TSS) was first described in Lancet (Todd et al., 1978). Toxic shock syndrome is an acute onset, multiorgan illness which resembles severe scarlet fever. The illness is caused by Staphylococcus aureus strains that express TSS toxin-1 (TSST-1), enterotoxin B, or enterotoxin C. TSST-1 is associated with menstrual TSS and approximately one-half of nonmenstrual cases; the other two toxins cause nonmenstrual cases, 47% and 3% respectively. The three toxins are expressed in culture medical under similar environmental conditions. These conditions may explain the association of certain tampons with menstrual TSS. Biochemically, the toxins are all relatively low molecular weight and fairly heat and protease stable. Enterotoxins B and C share nearly 50% sequence homology with streptococcal scarlet fever toxin A; they share no homology with TSST-1 despite sharing numerous biological properties (Schlievert and Blomster, 1983; Bohach et al., 1990) Numerous animal models for development of TSS have suggested mechanisms of toxin action, though the exact molecular action is not known. The toxins are all potent pyrogens, induce T lymphocyte proliferation, require interleukin-1 release from macrophages, suppress Ig production, enhance endotoxin shock, and enhance hypersensitivity skin reactions. The genetic control of the toxins has been studied and suggests the exotoxins are variable traits. Toxic shock syndrome is a serious, potentially lifethreatening condition resulting from an exotoxin of S. aureus. Presenting symptoms include high fever, diarrhea, nausea, and vomiting, progressing to hypotension, oliguria, conjunctival hyperemia, and an erythematous rash over the trunk, abdomen, and extremities. Most, but not all, cases of TSS have been associated with the use of tampons during menstruation, post-surgical infections, and stab wounds. It can occur in post-rhinoplasty patients with and without nasal packing. There have been cases reported in the literature of TSS associated with suction-assisted lipectomy (SAL). In 1978 Todd described seven children aged between 8 and 17 years who presented with a high fever, headache, confusion, diarrhea, and a scarlatiniform rash, some of whom developed prolonged shock, disseminated intravascular coagulation, renal failure, and hepatic abnormalities. One patient died and all suffered from fine desquamation of the skin of the hands and feet. A phase group I S. aureus was isolated from mucosal or sequestered (empyema, abscess) sites in five of the patients and this was found to produce a previous exotoxin which caused a positive Nikolsky sign in newborn mice (producing a cleavage in the basal layer of the

skin as opposed to the granular layer cleavage seen with the exfoliation produced by the phase group II staphylococcus). Todd also commented on the similarity between his cases and cases of Kawasaki’s disease, Rye’s syndrome, and staphylococcal scalded skin syndrome, and it was later suggested that reported cases of “adult Kawasaki’s disease” were probably cases of TSS. In 1980 Schrock was the first to describe an association between TSS and tampon usage in menstruating women. This association was confirmed by further studies. TSS was first linked (by the CDC) with menstruation in May 1980 and with Rely users in August and September of that same year. The manufacturer withdrew Rely from the market immediately. The first TSS sort trial was held in 1982. In a Wisconsin study of 38 cases, it was found that 37 were women, 35 of these cases occurred during menses, and 97% of these patients used tampons. The overall incidence of TSS in this study was found to be 6.2 cases; 100,000 menstruating women with a mortality rate of 2.6% compared with the national figure of 8.4%. The first case reported in the United Kingdom was in November 1980. Since then there have been over 80 cases reported to the Public Health Laboratory at Colindale. The disease predominantly affects menstruating women in the 20–30 year age group and the incidence seems to be very low in non-Caucasians. Before antibiotic treatment had been started, isolated S. aureus belonging to phase group II from 74% of vaginal and cervical culture had been obtained from TSS patients. In both these studies, blood cultures from TSS patients were negative; this finding is consistent with other reports suggesting, as Todd initially postulated, that the clinical features are due to the absorption of a toxin. However, there have been more recent reports where S. aureus has been isolated from blood cultures, although the incidence is very low. The strong association with tampon usage led to the investigation and comparison of brands with a special interest in the “superabsorbent” tampons containing carboxy methyl cellulose and polyester foam that were introduced in 198. No significant difference was found between types, although the Rely tampon (Proctor and Gamble) was withdrawn from the market in the United States. It was suggested that inherently contaminated tampons might be the cause, but absence of positive cultures for S. aureus and occurrence of TSS among users of all brands make this unlikely. Most attention has been made to tampon, mensesrelated TSS, although the original description of TSS was in children with S. aureus being isolated from various sites. Since then Reingold et al. (1982a, b) have made an extensive review of 130 cases of nonmenstrual TSS with S. aureus being isolated from a variety of sites including surgical wounds, adenitis, bursitis, postpartum infections, packed nasal cavities, and deep abscesses. Of the nonmenstrual TSS patients, the mortality rate was 9.2%, 68% were women, and the patients were significantly older. As with menses-related TSS, the incidence of bacteremia was low (3%). Recurrence of TSS occurs in the menstruating group only, with the incidence as high as 28%. The second episode usually occurs during the first or second menstrual period after the initial episode, which is usually the most severe. The incidence of recurrences can be reduced following treatment with the appropriate anti-staphylococcal antibiotics during the initial episode and cessation of tampon use in the subsequent periods. There were reported to be 1,138 cases of TSS, with 38 deaths, in 1980 and as many as 94 overall. Whether due to

Chapter 20:

a new, more pathogenic strain of bacteria, or to more absorbent tampons, or some contaminant in tampon materials (or some combination of these factors), the need to ensure that tampons neither seriously alter vaginal microflora nor promoted production of TSST-1 was clear, and (as described in Chapter 16) such testing is now performed prior to device approval. It is not clear what the relative contributing weights of the different components of the TSS problem were. Swift and decisive action before the public definitely served to reduce the public’s perception of the continuation of the problem.

LATEX ALLERGY The advent of public awareness and concern about acquired immune deficiency syndrome (AIDS) during the mid-1980s greatly increased the demand for condoms and gloves—two products of which the majority are made from latex. Protective gloves, commonly associated with aseptic surgical techniques, are now worn in virtually all procedures involving exposure to blood, body fluids, or mucous membranes. The AIDS epidemic and worry of infection transmittal have been an impetus for the donning of latex gloves by health-care professionals, equipment sales and service technicians, office cleaning staff, and individuals responsible for waste removal. Approximately 20 billion latex gloves were used in 1991 alone (Fay and Sullivan, 1992). As this utilization of gloves continues to escalate, increasing numbers of physicians, nurses, dentists, laboratory technicians, and patients are reporting latex hypersensitivity reactions in great frequency and severity. Delayed hypersensitivity reactions in the form of allergic eczematic contact dermatitis to natural rubber such as latex surgical gloves or other latex products has been observed for many years. Since 1979, when the first case of immediate hypersensitivity (contact urticaria) from latex was reported by J. K. Nutter (Gonzales, 1992) as a latex medical problem (Shumnes and Darby, 1984; Estlander et al., 1986). Between 1988 and 1992, the U.S. Food and Drug Administration (FDA) received reports of more than 1000 systemic type I allergic reactions to latex, of which at least 15 were fatal (Scarbeck, 1993). As a result of the regulatory requirements from the Centers for Disease Control (CDC), Occupational Safety and Health Administration (OSHA), and state regulatory boards, the wearing of gloves is no longer limited to health-care facilities. Workers in nursing homes, schools, law enforcement agencies, reference laboratories, emergency medical services, and linen services, along with hospitals and funeral services, are donning latex gloves. Due to the increased use of latex, latex hypersensitivity affects not only the surgical team, but these high-risk groups as well. Fisher (1991) reported 7% of physicians, nurses, and dentists now acquire a delayed allergic, eczematous contact dermatitis, whereas 3% show an immediate allergic urticarial reaction to the aqueous protein fraction of latex. This urticarial reaction may be accompanied by anaphylaxis. Iatrogenic and intraoperative contact urticaria and anaphylaxis of patients contacting rubber gloves worn by physicians are being reported with more frequency. Due to the increase in these reports, in March 1991, the FDA alerted health-care professionals about the potential of severe allergic reactions to medial devices made of latex.

Case Histories and Problem Resolution

279

Presently, the incidence of latex sensitization is unknown. Latex sensitivity has been reported in one of 800 patients (0.128%) prior to surgery (Turjanma, 1993). However, latex allergy in children with spina bifida ranges from 28 to 67% (Kelly et al., 1994). Health-care workers with atopic allergy have been reported to have a 24% prevalence of a positive latex skinprick test, while the overall prevalence in health-care workers is 7–10% (Arellano et al., 1992). Latex is very widely used—not just in many medial devices (gloves, condoms, catheters, and the stoppers of syringes, for example), but also in many everyday used items (balloons and tires, for example). Once an individual is sensitized, exposure to any of these can evoke a potentially deadly response. There are at least two operative mechanisms for the range of responses seen (Lang, 1996).

Delayed hypersensitivity Type IV delayed hypersensitivity is the predominant immunological response (82%) to natural rubber latex (Sussman, 1995). Although skin irritation may be due to friction, maceration, frequent hand washing, and antimicrobial soap, exposure to rubber gloves or glove powder should be considered a potential allergen. The usual delayed hypersensitivity reaction is induced by low-molecular-weight chemical additives: vulcanizers, stabilizers, accelerators (thiurams, mercaptobenzothiazole, carbamates), and antioxidants in p-paraphenylenediamine black-rubber mix (Fay and Sullivan, 1992; Sussman et al., 1991). The immunologic pathways and mediators to delayed contact dermatitis include the development of immunized T lymphocytes able to recognize the antigen that usually occurs in seven days. It is believed that the Langerhans cell mediates this process by engulfing the antigen on primary exposure, migrating to lymph nodes, and then presenting the antigen to helper T cells. The sensitized helper T cell then reenters the skin immunologically prepared for the next encounter with the antigen. Upon secondary exposure to the antigen, the lymphocytes release cytokines that recruit macrophages and other inflammatory cells to elicit the cutaneous reaction that will peak in approximately 48 hours (Gonzales, 1992).

Immediate Hypersensitivity Immediate hypersensitivity reactions tend to be more serious because they involve not only the skin but also mucosal surfaces. These manifestations occur within minutes and may include contact urticaria, angioedema, rhinitis, and respiratory symptoms such as dyspnea and asthmatic attacks. Systemic reactions have been reported to occur in patients with latex protein allergies and are more likely to occur with prior mucosal exposure to gloves, balloons, or dental dams. Patients subject to immediate reactions are at risk of severe or fatal consequences if the hypersensitivity is unrecognized when they are examined or operated on by health care professionals wearing latex gloves. The immunological pathway or mechanism involved in immediate latex hypersensitivity is quite different from that of the delayed reactions. Immediate latex hypersensitivity reactions are believed to be mediated by immunoglobulin E (IgE) (Gonzales, 1992; Slater, 1992; Sussman et al., 1991). Interaction of IgE and antigen induces local release of histamine and arachidonic acid metabolites. The release of these agents results in increased vascular permeability to the antigen and systemic immediate reactions. It is believed that the antigen in the IgE-mediated response is a latex protein or polypeptide

280

Safety Evaluation of Medical Devices

(Warpinski et al., 1991). It is not clear at all, however, that the increased incidence and severity of responses seen since the demand for gloves and condoms went up is due solely to there being increased numbers of people being exposed with increased frequency. Certainly such is not the case with the most sensitive population, the children with spina bifida. A significant component in the causality may well be a change in either sources of latex or processing procedures (or both). A spectrum of IgE-mediated allergic responses to latex was presented in an article by Sussman and co-workers (1991). They reported 14 patients sensitized by exposure to latex gloves. The manifestations in response to the latex exposure varied greatly according to the route of latex antigen presentation. However, the patients’ symptoms often occurred immediately after exposure to latex. Skin contact usually elicited urticaria. Allergic rhinitis, conjunctivitis, and asthma were usually the result of exposure to airborne particles of latex. Due to contact with surgeons’ gloves intra-operatively, systemic effects from latex occurred. Anaphylactic shock included these symptoms plus tachycardia and hypotension. Sussman et al. (1991) reported positive latex skin tests from all the patients. Serum IgE antibody to latex was found with latex radioallergosorbent test (RAST) in all patients tested except one. Other studies have shown immediate hypersensitivity reactions to latex to be IgE-mediated by skin testing, basophil histamine release, RAST, enzyme-linked immunosorbent assay (ELISA), and IgE immunoblots. Though traditional in vivo delayed contact sensitization testing should have been sufficient to detect the increased delayed sensitization response potential of latex products, it did not. More importantly, such traditional sensitization tests cannot function to detect IgE-based responses (though the mouse PLNA would). More extensive immunotoxicity testing is clearly called for in the case of a natural product-based, high-exposure material used for devices. And careful attention must be paid to source materials provided by vendors for approved devices, particularly when demands for the material are sharply increased over a brief period of time.

SILICONES IN DEVICES As presented back in Chapter 2, silicones have come into broad use in medical devices since the 1950s (McGregor, 1953). In various chemical and physical forms, they are present in syringes (as lubricants), in shunts and catheters (in tubing), and in implants (as films and gels). Since 1964, silicone-filled breast implants have been put into an estimated one million women, with the average cost of the actual implant being $195 in 1990. Prior to this, there were some claims of association of breast cancer with implants and speculative reports of immunologically mediated health problems associated with exposure to silicone in implants and from kidney dialysis (Laohapand et al., 1982; Baldwin and Kaplan, 1983; Bommer et al., 1983), but these had neither been scientifically verified (Heggers et al., 1983) nor successful in litigation in court. A persistent problem was (and remains) a confusion on the part of some as to the difference between silica (of which some forms are immunotoxic; Uber and McReynolds, 1982) and the silicones. Starting in early 1991, there was an explosion in litigation over silicone breast implants with suits by women claiming harm from the implants. Small amounts of silicone fluid from the gel have leaked through the outer envelope of

the implant. (This leakage was not noticeable to the women and tended to be contained within the capsule of scar tissue that inevitably forms around the implants.) In some women, the capsule of scar tissue contracted excessively, distorting and hardening the breasts and often causing discomfort. In about 5% of women, an implant ruptured, releasing silicone gel into the surrounding tissues and flattening the breast. But these local complications, unpleasant as they were, were not the basis for most of the alarm about breast implants, nor were they the focus of the multimillion-dollar lawsuits. Instead, a growing number of Americans had come to believe that breast implants could cause devastating effects on the rest of the body. In particular, silicone gel-filled implants were said to be responsible for an ill-defined constellation of disorders known as connective tissue diseases. These diseases—which include systemic lupus (SLE), rheumatoid arthritis, and scleroderma—are thought to involve a disturbance in the immune system that turns the body’s protective defenses against itself. The result is an autoimmune disease that can produce profound weakness and fatigue along with variable damage to the joints, skin, and internal organs. The symptoms and ills identified have continued to shift and such poor definition of a syndrome has caused significant uncertainty as to mechanism (indeed, as to the actual existence of a silicone-related disease moiety at all). This makes performance of an epidemiological study capable of deciding the issue impossible. One theory was that silicone, leaking slowly from the implants, provokes an immune reaction that then somehow turns into an autoimmune process. It was the unproven theory (and variations on it) that served as the basis for the largest lawsuits. But the question of whether or not silicone gel-filled breast implants cause connective tissue disease is not ultimately a matter of opinion or legal argument; it is a matter of biological fact. Either they cause connective tissue disease (alone or in conjunction with other factors), or they do not. And the only way to answer the question is through epidemiologic studies. To be sure, there are many individual stories (some medically proven, others not) of connective tissue disease developing after the placement of breast implants, but these anecdotal reports alone do not constitute evidence that the implants caused the disease. One such case is dramatically presented in the book Informed Consent (Byrne, 1996), though not in a scientifically critical or dispassionate manner. They could well represent pure coincidence. Each year, for example, about 100 women with breast implants can be expected to come down with lupus or scleroderma by chance alone. Since connective tissue disease can occur in women with or without implants, the only way to demonstrate that implants actually contribute to the disease is to show that the incidence is significantly higher in women with breast implants than in those without. Yet it was not until June 16, 1994, two years after breast implants were taken off the market and two months after the class-action settlement was announced, that the first such study of the possible link between breast implants and connective tissue disease was published. And that study failed to find a link. In the study, May Clinic researchers compared a group of 749 women who had received breast implants between 1964 and 1991 with 1498 of their neighbors, matched for age. The researchers found that the implant group was no more likely to develop connective tissue disease (or related symptoms and abnormal tests) than the

Chapter 20:

group without implants. This was only one study, of course, and was not large enough to rule out some increase in risk. But it did cast doubt on the link between breast implants and connective tissue disease at a time when many people assumed the theory had been proven. Meanwhile, at least four other well-designed epidemiologic studies were underway. The largest was a retrospective cohort study of about 450,000 American women in the health professions. Although they have not been published yet, interim results indicate no association between breast implants and connective tissue diseases, with the possible exception of rheumatoid arthritis. Another large retrospective cohort study, published in the New England Journal of Medicine in June 1995, also found no association between implants and connective tissue diseases, and so addressed the claim that disease caused by breast implants may not fulfill all the usual criteria for “classical” connective tissue disease. The remaining two epidemiologic studies, published in 1994 and 1995, dealt just with scleroderma. Scleroderma, a disease characterized by extensive scarring of the skin and sometimes of internal organs, is the connective tissue disease most closely linked by anecdotal reports to silicone breast implants. Neither study could find an association. While all together these studies do not mean that there can be no link between connective tissue disease and breast implants, they do mean that any risk of connective tissue disease from implants is so small that it has been impossible to detect. The most recent epidemiology study (Hennepens et al., 1996, released in February), did find “a small but statistically significant increased risk” (one in 30,000 per year implant exposure) of contracting immune-system illness. This small potential difference of risk, however, may only represent the well-recognized influence of self-reporting in a highly publicized situation. Wong (1996) subsequently overviewed and performed a meta-analysis on 15 epidemiology studies that he had identified on breast implants and connective tissue disease, and found that they “did not provide any evidence of a causal relationship between silicone breast implants and connective tissue disease.” The result of litigation to date, however, is sobering. Facing one product liability case after another, the breast implant manufacturers, while maintaining that the implants were safe, eventually agreed to the largest class-action settlement in the annals of American law. In April 1994, the major manufacturers settled for $4.25 billion, a billion of which was explicitly set aside for the lawyers involved. Although nearly any woman with implants would be entitled to something under the terms of the settlement, women were permitted to opt out if they thought they could do better on their own. As of June 1, 1995, some 440,000 women (about a quarter of all women with breast implants in this country) had registered to participate in the class-action settlement but, with vigorous encouragement form plaintiffs’ attorneys, about 15,000 (half of them foreigners, for whom the terms of the settlement are less generous) have opted to seek higher damages individually. The terms of the class settlement were remarkably generous and broad. All women with breast implants were entitled to compensation if they had, or within the next 30 years developed, any of ten connective tissue diseases or symptoms suggestive of such disease, provided the symptoms began or worsened after the implants were placed. Of the $4.25 billion, $1.2 billion was set aside for women claiming to have implant-related illnesses already—248,500 women

Case Histories and Problem Resolution

281

of the 440,000 registered as of June 1. The amount of compensation was to be determined by the type of disorder, its severity, and the woman’s age at onset. A chart or grid sent to all women with breast implants showed the exact amounts. For example, a woman over 56 with mild Sögren’s syndrome (dryness of the eyes and salivary glands) would receive $140,000. Claimants were not required to show that the implants were related to the illness. In addition, women who were not ill could receive lesser amounts for emotional distress. They would also be reimbursed for all uninsured medical costs related to breast implants, including evaluations, treatment of implant rupture, and removal of implants. Husbands, other “significant others”, and children born before April 1, 1994, were also entitled to make claims. Children, for example, could claim compensation for injuries caused by their mothers’ implants (a particularly mysterious provision, since no such injury has ever been demonstrated). Women claiming current illness were required to submit substantiating medical records. If these were not sufficient to place the woman in the appropriate category on the grid, the woman’s doctor was to send the diagnosis, along with copies of relevant records. Beyond this, there would be no attempt to verify the woman’s medical condition. A doctor’s diagnosis or the medical records would be challenged only if they failed to meet the eligibility requirements (such as swollen joints or abnormal substances in the blood) which can be objectively measured; some (such as fatigue or muscle aches) cannot. In fact, it would be possible to qualify for compensation without any objective manifestations of illness whatsoever. For example, a woman could claim joint and muscle aches, disturbed sleep, fatigue, and burning pain in the chest, none of which can be objectively verified by her doctor or anyone else and collect up to $700,000. In the end, the class-action settlement has seemingly unraveled. In early May of 1995, Judge Sam C. Pointer, the Alabama federal judge who was overseeing the settlement, announced that the size of the compensation for each woman would have to be revised downward. It was apparent that the $1.2 billion set aside for current claims would nowhere near cover them. And, as it looked like each woman will receive less, more women have opted out of the settlement and gone for jury trials, leaving the manufacturers little reason to stay in. A desperate Judge Pointer charged the attorneys representing both sides to try to negotiate a new agreement by August 30 1995, presumably one involving increased contributions from the manufacturers. “There are just too many sick women,” said Ralph Knowles, one of the lawyers representing the plaintiffs. “I didn’t think it was going to be anything like that. If I did, we would have never agreed to the $4.25 billion.” But, whatever happens to this settlement, a great deal of damage has already been done. It did not take plaintiffs’ attorneys long to realize that breast implants are not the only medical devices on the market that contain silicone. Already, several class actions have been filed on behalf of the approximately 300,000 men with silicone -containing penile implants. Breast implant litigation has also contributed to new alarmism about Norplant, the highly reliable contraceptive that is placed under the skin of the arm in six very small, silicone-coated rods. The number of product liability lawsuits against Wyeth-Ayerst, the manufacturer of Norplant, and its parent company, American Home Products, has, according to a New York Times story, swelled from twenty in Norplant’s first three years on the market to 180 in 1994. Forty-six class-action

282

Safety Evaluation of Medical Devices

suits have been filed on behalf of Norplant users, and implants of the contraceptive have fallen from 800 a day to sixty. The illnesses attributed to it include autoimmune and connective tissue-like disorders. Yet, as with breast implants, there is no evidence to implicate Norplant in these disorders. It should be noted however, that even in the 1990s, despite vigorous attempts (including extensive advertising for clients) plantiffs’ attorneys were unable to mount successful litigation attacks on other large use devices that had silicone in them (such as diabetics, whose multiple daily injections each included a small amount of silicone lubricant from the syringe). More worrisome still is the indirect threat to all medical devices—whether they contain silicone or not. Under our liability laws, plaintiffs can make claims against any party involved in the manufacture of an allegedly harmful product, no matter how remote the involvement. Suppliers of raw materials (or biomaterials) for medical devices can be sued, then, even if they have nothing to do with the design and manufacture of the product. The sale of biomaterials for medical devices is a small part of the business of most big suppliers, so the resulting revenues can quickly be offset by the legal liabilities. Many large suppliers of biomaterials pulled back from the market, or have attempted to severely limit the conditions of use of their materials in even the upstream use in devices. Dow Corning, a supplier as well as a manufacturer, has drastically scaled back sales of silicone to other manufacturers of medical devices and may stop selling it altogether. The embargo will probably affect a wide variety of silicone-containing devices, ranging from the useful to the vital. Among them are cardiac pacemaker wires, artificial joints, mechanical hear valves, intraocular lenses (used after cataract surgery), implantable arteriovenous shunts for people on chronic dialysis, and shunts for people with hydrocephalus (a potentially lethal condition in which fluid accumulates in the brain). Dow Chemical Company has stopped supplying a material used in pacemaker components. DuPont announced in 1993 that it would sever connections with the permanent medical implant industry. It will no longer provide medical manufacturers with Dacron polyester, which is used in vascular grafts, or a number of other materials. In DuPont’s calculus, what had happened with silicone could happen with any other constituent of medical devices. Meanwhile, an extensive battery of immune toxicity tests has been performed on the silicones that were and are used in implants. To date, none of these have established any immune system effects. And the USFDA allowed the return of silicone breast implants to the market. Under these conditions, a large number of important medical products may become scarce or even unavailable. In May 1994, Senator Joseph Lieberman of Connecticut, then the chairman of the Governmental Affairs Subcommittee on Regulation and Government Information, held hearings on the impact of product liability suits on the availability of medical devices. Among those who testified was Elenor Gackstatter, president of Meadox Medicals, a manufacturer of vascular grafts and other devices. Gackstatter said that she had tried to contact fifteen alternative suppliers of polyester yarn after Du Pont announced it would no longer supply Dacron to her company. None of them, even foreign suppliers, would deal with American manufacturers because of the liability risks. Many manufacturers have a two- or three-year supply of biomaterials on hand, but when that is depleted, there may be a serious shortage. Said Lieberman, “This is a public health time bomb, and the lives of real people are going to be lost if it explodes.”

In 2007 and 2008, federal courts (going all the way to the US Supreme Court, in Riegel vs. Medtronics) have held that for medical devices and their components, USFDA approval acts as a bar (or shield) against product safety litigation unless the approval was based on knowingly false or incomplete data Any medical device or pharmaceutical with a sufficiently large and long patient exposure base will have the potential of evoking unforeseen adverse responses in one or more subpopulations of the exposed patient population, either in reality or in the perception of the public and civil legal system. Such devices or agents must be fully tested against the range of possible biologic effects, and mere adherence to existing guidelines (with careful and informed scientific consideration) is not adequate. Perceived subpopulation “risks” must be aggressively investigated to fully understand both mechanisms and the relative extent of the risk.

PROBLEM SOLVING When the biological evaluation of medical devices is approached as a routine function involving nothing more than successfully passing a series of biocompatibility tests, there is little opportunity for innovatively managing positive test results, which indicate that a material is toxic. But blind compliance is not the intent of the International Organization for Standardization (ISO) or FDA. Manufacturers have the freedom, and the responsibility, to apply the ISO biological evaluation standard in a way that ensures the biological safety of their devices while conservatively managing resources. Neither ISO 10993-1 nor the FDA memorandum on its use specify pass/fail criteria for biological testing, recognizing that it is almost impossible to set general criteria and that manufacturers are in the best position to determine what level of toxicity is acceptable for their products. Based on careful comparisons of biocompatibility and clinical data, some companies have determined the highest safety test score relating to unacceptable performance for a specific class of products and use this value as the pass/fail criterion. At many companies, however, there is a tendency to panic when biological safety test results are positive. The possibility of a positive result can call into question carefully considered material choices, threaten costly delays in product development schedules, and raise doubts about test strategies. Manufacturers may hastily pursue many directions at once, ending with an array of conflicting information. Or they may forget that the goal of biological safety testing is to determine whether a material or device is safe for its intended use, and not (necessarily) to determine the cause of toxicity. The best approach to any situation where the ideal may be unobtainable is to follow a planned course of action, first confirming the facts at hand, then considering the options for future actions. Applied to the problem of positive biocompatibility test results, this means methodically confirming that the test procedure was followed as intended and that the test result is reproducible, and then considering whether the toxicity can be eliminated or is acceptable. The steps in this procedure are (Stark, 1996): Confirming the test procedure. a. Was the test performed and interpreted properly? b. Was the proper test article employed (and so identified)?

Chapter 20:

c. Was the test article properly sampled and prepared (including storage, extraction and any manipulation)? d. Was the test properly conducted? Confirming that the proper material was supplied. a. Was the material properly manufactured? b. Sterilized? c. Cleaned? Confirming the reproducibility of the test. a. Can the results be replicated? Eliminating the source of toxicity. If all the above are confirmed, can the cause of test failure/toxicity be removed from the product? a. By using different materials or formulations? b. By washing, rinsing, or otherwise removing a process contaminant? c. By improving process or material quality control? If all of these steps do not lead to a product which passes the required test paradigm, then one must consider whether the identified toxicity presents an unacceptable risk, or whether the hazards are either justifiable or irrelevant to the actual use of the device. While the ideal situation is eliminating any potential toxicity, this is not always possible. Toxicity may be intrinsic to the product and impossible to eliminate without compromising product function. One familiar example is the electronic componentry of pacemakers and cochlear implants. The toxic circuitry in these devices must be contained within a nontoxic case so it cannot leach out and injure the implant recipient. Another example is a simple product called an adhesive remover. No matter how it is formulated, the product is always a mixture of organic solvents that carry with them the possibility of systemic toxicity subsequent to skin absorption. Medical devices that must cure in situ are also intrinsically toxic. The curing process of products such as casts, dental cements and bone cements may involve the generation of free radicals or other reactive chemical moieties, or may be exothermic. In addition, implants made from nickel alloys carry an intrinsic level of toxicity. Nickel is a cardiac toxin, an oxytoxic agent, and a common sensitizing agent (an estimated 5% of the population are allergic to nickel contact). The possibility of nickel being released into the biological environment always poses the risk of toxic response. There are also some devices whose functions result in injury—for example, a medical tape designed to hold an appliance onto the skin. If the appliance is life-supporting, the tape will be expected to adhere to the skin with some high degree of tenacity so that the device will not fall away. This high adhesion level is likely to result in skin injury when the tape is ultimately removed. In each of these examples, the toxicity is intrinsic to the device; suitable (nontoxic) alternative materials do not exist, or the device will not function as intended if the injurious material is removed. The manufacturer is left with no alternative but to accept the toxic material. The strategy becomes one of justifying its use.

Justifying Use of the Material There are three approaches to justifying the use of a toxic material in a medical device. The first is to compare the level of toxicity of the material to a comparable material

Case Histories and Problem Resolution

283

that is currently being used by the manufacturer. If the new material has a lower level of toxicity than the current one and the current one has a safe history of use in the marketplace, the use of the new material may be justified because it is a move in the direction of decreased toxicity and increased biological safety. The second approach is to compare the level of toxicity of the material to a comparable material that is currently being used in a competitive product. Again, if the new material has a lower level of toxicity than the competitive material in the same biological safety test, and the competitive material has the safe history of use in the marketplace, the use of the new material may be justified. In the third approach, the maximum dose and the no-observable-adverse-effect level (NOAEL) for the material are calculated and then compared. To determine the level at which no adverse effect occurs, the sample is titrated by using decreasing amounts in the test system. The highest concentration of sample at which no effect is observed is the NOAEL, which can be expressed in units, surface area, weight, or volume of material. The maximum dose of a material equals the units, surface area, weight, or volume of material to which a patient will be exposed during a typical course of therapy. If a material’s maximum dose is 100-fold less than its NOAEL, the material is considered safe for use. (The 100-fold criterion is based on a 10-fold variation between species and a 10-fold variation within species.) The comparison must be repeated for each biological safety test giving a positive response. The NOAEL approach has been employed to justify the use of nickel alloys in implants. The amount of nickel released by in situ corrosion is compared with the maximum permissible amount of nickel that can be given per day in intravenous fluids, which was determined from intravenous injection of nickel in dogs. If the release of nickel through corrosion is less than the amount that can be safely given in intravenous fluids, the alloy would be considered to be safe for implantation. For such situations, one must conduct (and provide to regulatory bodies) a risk assessment.

RISK ASSESSMENTS ISO 14971-1 (2000) sets forth the requirements and means of conducting and presenting a risk assessment (analysis) for medical devices. While every medical manufacturer desires to construct its devices from entirely safe materials, in reality not all materials are entirely safe. Generally, if one looks long enough at small enough quantities, some type of risk can be associated with every material. Risk can be defined as the possibility of harm or loss. Health risk, of course, is the possibility of an adverse effect on one’s health. Risk is sometimes quantified by multiplying the severity of an event times the probability the event will occur, so that: Risk = Severity × Probability While this equation appears useful in theory, in practice it is difficult to apply to the biological safety of medical devices. The process known as health-based risk assessment attempts to provide an alternative strategy for placing health risks in perspective.

284

Safety Evaluation of Medical Devices

A structure for the risk assessment process has been detailed in a publication prepared by the U.S. National Academy of Sciences (NRC, 1983; Gad, 1999). Although devised primarily for cancer risk assessment, many of the provisions also apply to the assessment of other health effects. The major components of the paradigm are: (1) hazard identification, (2) dosage–response assessment, (3) exposure assessment, and (4) risk characterization. This general approach to risk assessment was adapted to medical devices via the draft CEN standard Risk Analysis, published in 1993, and more recently via the ISO standard, ISO 14538—Method for the Establishment of Allowable Limits for Residues in Medical Devices Using Health-Based Risk Assessment (ISO, 1998). At the present time, FDA is also working to develop a health-based risk assessment protocol adapted to medical devices. Informally called the Medical Device Paradigm, the document is not yet generally available. Some manufacturers may object that regulators are once again attempting to impose a “drug model” on medical devices. However, we shall see in the following pages that judicious application of these risk assessment principles can provide a justification for using materials that carry with them some element of risk, and that may, under traditional biocompatibility testing regimes, be difficult to evaluate or be deemed unsuitable for medical device applications.

Method Hazard Identification. The first step in the risk assessment process is to identify the possible hazards that may be presented by a material. This is accomplished by determining whether a compound, an extract of the material, or the material itself produces adverse effects, and by identifying the nature of those effects. Adverse effects are identified either through a review of the literature or through actual biological safety testing. Dose–Response Assessment. The second step is to determine the dose response of the material—that is, what is the highest weight or concentration of the material that will not cause an effect? This upper limit is called the allowable limit. There are numerous sources in the literature of data from which to determine allowable limits; some will be more applicable than others, and some may require correction factors. Exposure Assessment. The third step is to determine the exposure assessment by quantifying the available dose of the chemical residues that will be received by the patient. This is readily done by estimating the number of devices to which a patient is likely to be exposed in a sequential period of use (for instance, during a hospital stay) or over a lifetime. For example, a patient might be exposed to 100 skin staples following a surgical procedure, or to two heart valves in a lifetime; thus, the amount of residue available on 100 skin staples or two heart valves would be determined. Risk Characterization. Characterizing the risk constitutes the final step of the process. The allowable limit is compared with the estimated exposure; if the allowable limit is greater than the estimated exposure by a comfortable safety margin, the likelihood of an adverse event occurring in an exposed population is small, and the material may be used.

Nitinol Implant Nitinol is an unusual alloy of nickel and titanium that features the useful property of “shape memory.” A nitinol part can be given a particular shape at a high temperature, then cooled to a low temperature and compressed into some other shape; the compressed part will subsequently deploy to its original shape at a predetermined transition temperature. This device gives a transition temperature at approximately the temperature of the body (37°C). Hazard Identification. One concern with using nitinol in an implant application is the potential release of nickel into the body. Although nickel is a dietary requirement, it is also highly toxic. It is known to cause dermatitis, cancer subsequent to inhalation, and acute pneumonitis from inhalation of nickel carbonyl, and to exert a toxic feature is particularly beneficial for vascular implant applications in which the shape of the device in its compressed state eases the insertion process. The nitinol deploys as it is warmed by the surrounding tissue, expanding to take on the desired shape of a stent, filter, or other device. The transition temperature depends on the alloy’s relative concentrations of nickel and titanium; a typical nickel concentration of 55–60% is used in medical effects on cellular reproduction. It is a known sensitizer, with approximately 5% of the domestic population allergic to this common metal, probably through exposure from costume jewelry and clothing snaps. The biocompatibility question at hand is whether or not in vivo corrosion of nitinol releases unsafe levels of nickel. Dose–Response Assessment. A search of the world medical literature revealed that the recommended safe level of exposure to nickel in intravenous fluids is a maximum of 35 µg/day. This value can be taken as an allowable limit of nickel exposure for a 70-kg (154-lb so-called “standard male”) adult. The intravenous fluid data are based on subjects that are compatible to the patients who will be receiving nitinol implants. The data are for humans (not animals), for ill patients (not healthy workers or volunteers), and for similar routes of exposure (intravenous fluid and tissue contact). For these reasons, no safety correction factor need be applied to the allowable limit of exposure. Exposure Assessment. The available dose of nickel from nitinol implants can be estimated from data found in the literature. In one study, dental arch wires of nitinol were extracted in artificial saliva, and the concentration of nickel measured in the supernatant. Corrosions reached a peak at day 7, then declined steadily thereafter. The average rate of corrosion under these conditions was 12.8 µg/day . cm2 over the first 28 days. Risk Characterization. A comparison with the available dose with the allowable limit for intravenous fluid levels shows that there is approximately a three-fold safety margin, assuming that the implanted device is a full 1 cm2 in surface area. (Devices with less surface area will contribute even less to the nickel concentration and have an even larger safety margin). Considering the high quality of the data, a threefold safety margin is sufficient to justify using nitinol in vascular implants.

Wound-Dressing Formulation CASE STUDIES We can best get a sense of how these standards work by looking at some actual case studies that illustrate the risk assessment process.

Today’s wound dressings are highly engineered products, designed to maintain the moisture content and osomotic balance of the wound bed so as to promote optimum conditions for wound healing. Complex constructions of hydrocolloids and superabsorbers, these dressings are sometimes used

Chapter 20:

in direct tissue contact over full-thickness wounds that penetrate the skin layers. Hazard Identification. There have been reports in the literature of patients succumbing to cardiac arrest from potassium overload, with the wound dressing as one of the important contributors of excess potassium in the bloodstream. The effects of potassium on cardiac function are well characterized. Normal serum levels of potassium are 3.8–4 milliequivalents per liter. As the potassium concentration rises to 5–7 mEq/L, a patient can undergo cardiac arrest and die. The biocompatibility issue to be explored is whether or not a wound dressing formulation might release dangerous levels of potassium if used on full-thickness wounds. Dose–Response Assessment. An increase of approximately 1 mEq/L of potassium is likely to provoke mild adverse events in most patients. Assuming that the average person’s blood volume is 5 L, a one time dose of 5 mEq/L of potassium may begin to cause adverse reactions. This value can be considered to be the allowable limit of potassium for most patients. Exposure Assessment. Let us suppose that each dressing contains 2.5 g of potassium bicarbonate. Since the molecular weight of potassium bicarbonate is 100 g/mole, each dressing contains 0.025 mole of sodium bicarbonate, or 0.025 mEq/L of potassium. If a patient were to use four dressings in a day, the available dose of potassium would be 0.1 mEq/day. Risk Characterization. Comparing the available dose of potassium (0.1mEq) to the allowable limit (5 mEq) shows that there is a 50-fold safety margin. Considering that patients may be small in size, or may receive potassium from additional sources such as intravenous fluids, this safety margin is too small, and so the dressing should be reformulated.

Perchloroethylene Solvent A manufacturer of metal fabricated parts uses perchloroethylene to clean the finished pieces. Perchloroethylene has many advantages as a cleaner and degreaser: it is highly volatile, does not damage the ozone layer, and is very effective as a precision cleaning solvent. The most common use of perchloroethylene is in the dry cleaning industry, but it is also commonly used in the electronics industry to clean circuit boards.. Hazard Identification. The downside of perchloroethylenes is that it is highly toxic, with a material safety data sheet, several pages in length, listing adverse effects ranging from dizziness to death. Biocompatibility testing on solvent-cleaned parts would be meaningless; the solvent concentration on the part is so small that any effects of the solvent would be masked by the natural biological process of the test animals. The biocompatibility question that must be answered is whether or not sufficient residual perchloroethylene remains on the cleaned metal parts to pose a health hazard. Dose–Response Assessment. Threshold limit values (TLVs) are values that indicate the maximum level of a chemical that a healthy worker could take in on a daily basis over the course of his or her work life without experiencing any adverse effects. The TLV for perchloroethylene is 50 ppm/day (50 mL of perchloroethylene per 103 L of air) by inhalation. The average person inhales 12,960 L of air per day, making this equivalent to 650 mL of perchloroethylene per day. Since the vapor density of perchloroethylene is 5.76 g/L, the TLV is equal to 2.7 g of perchloroethylene per day by inhalation. Because TLVs for inhalation—as opposed to direct tissue exposure—are determined based on healthy

Case Histories and Problem Resolution

285

individuals (not ill patients), we will divide the TLV by an uncertainty factor of 100, i.e., 10 to account for a different route of exposure and 10 to account for healthy-to-ill persons. By this method, we obtain an allowable perchloroethylene limit of 37 mg/day. Exposure Assessment. To calculate an available dose of perchloroethylene, we need some additional information. In this case, the manufacturer brought a number of cleaned metal pieces into equilibrium within a closed jar then analyzed the headspace above the pieces by using high-pressure liquid chromatography to determine the concentration of perchloroethylene released. The concentration of perchloroethylene was undetectable by high-performance liquid chromatography. Since the limits of this analytical method are 2 ppb, this value was taken as the concentration of perchloroethylene in the headspace. Taking the weight of the metal pieces, the number of pieces tested, and the volume of the headspace, it was calculated that the amount of perchloroethylene per single piece was a maximum of 1.0 ng/piece. If we suppose that a patient might be exposed to a maximum of 50 pieces over a lifetime, then the maximum available dose of perchloroethylene from the pieces would be 50 ng. Risk Characterization. A comparison of the available dose (50 ng) to the allowable limit (37 mg/day) indicates an ample safety margin.

Ligature Material A manufacturer purchases commercial, black fishing line to use as a ligature in a circumcision kit. Because the ligature is not “medical grade,” a cytotoxicity test is routinely conducted as an incoming inspection test. It was assumed that a negative cytotoxicity test would be associated with an acceptable incidence of skin irritation. Hazard Identification. A newly received lot of the fishing line failed the cytotoxicity test. The extraction ratio of this material—of indeterminate surface area—was 0.2 g/mL, with a 0.1-mL aliquot of sample extract being applied to a culture dish. Thus, 0.2 g/mL × 0.1 mL = 0.02 g represents a toxic dose of fishing line. Dose–Response Assessment. A titration curve was obtained on the sample extract. If the sample was diluted 1:2, the test was still positive; however, if the sample was diluted 1:4, the test was negative. Thus, 0.02 g/4 = 0.005 g of fishing line, the maximum dose that is not cytotoxic. This value was called the allowable limit of fishing line. Exposure Assessment. Each circumcision kit contained about 12 in. of line, but only about 4 in. of the material was ever in contact with the patient. Since an 8-yd line was determined to weigh 5 g, the available dose of fishing line was calculated to be 5 g/288 in. × 4 in. = 0.07 g. Risk Characterization. A comparison of the available dose (0.07) with the allowable limit (0.005 g) convinced the manufacturer to reject the lot of fishing line.

Sources of Data Data for calculating the allowable limit of exposure to a material can come from many sources, most of them promulgated by industrial and environmental hygienists and related agencies. Review Chapter 2 on data sources for a better understanding of this area.

Uncertainty Factors An uncertainty factor is a correction that is made to the value used to calculate an allowable limit. It is based on the

286

Safety Evaluation of Medical Devices

uncertainty that exists in the applicability of the data to actual exposure conditions. Typically, uncertainty factors range in value from 1 to 10. For example, a correction factor of 10 might be applied for data obtained in animals rather than humans, or to allow for a different route of exposure. In other words, for every property of available data that is different from the actual application, a correction factor of between 1 and 10 is applied. If our first example had been of a small amount of data obtained in animals by a different route of exposure, an uncertainty factor of 1000 might be applied.

Safety Margins A safety margin is the difference or ratio between the allowable limit (after correction by the uncertainty factor) and the available dose. How large does a safety margin need to be? Generally, a safety margin of 100× or more is desirable, but this can depend on the severity of the risk under consideration, the type of product, the business risk to the company, and the potential benefits of product use.

CONCLUSIONS Ensuring the safety of medical devices (both new and, as a quality control concern, established) is an ongoing challenge. Strict adherence to regulatory requirements in the intended markets is essential but not sufficient. Careful and continuous consideration must be given to both the “known” potential problems (as presented in this text, for example) and also to those new areas of concern made visible by the continued advancement of science and extension of human experience. Immunotoxicity is currently the most apparent area of real and new but ill-defined concern. Medical device manufacturers have two predominant questions when it comes to material biocompatibility. The first is: “We have a material that we absolutely must use in our device, but it fails a biocompatibility test. Can we justify using the material anyway?” The second is: “We have a material that we absolutely must use in our device, but carcinogenicity and/or chronic toxicity testing are required. Can we justify omitting these tests?” Judicious application of health-based risk assessments can help with both of these issues, often providing a fast, cost-effective answer to both questions. Additionally, the seemingly safe and fixed field of regulatory safety assessment for devices is (and will remain through the current decade) in flux. The author believes that FDA (and other regulatory) requirements will continue to become more stringent until they are indistinguishable from those for pharmaceuticals. And the economic forces causing companies to have to compete in a world wide market will only intensify, making the regulatory standards for device approval confused (though harmonization under ICH will continue to make progress) and equivalent to the highest national standard in which companies intend to compete.

REFERENCES Arellano, R., Bradley, J. and Sussman, F. (1992). Prevalence of latex sensitization among hospital physicians occupationally exposed to latex gloves. Anesthesiology, 77: 905–908. Autian, J. (1972). Toxicity and health threats of phthalate esters. Review of the literature. Toxicology Information Response Center,

Article No. ORNC-TRIC-72-2, Oak Ridge National Laboratory, Oak Ridge, TN. Baldwin, C.M. Jr. and Kaplan, E.N. (1983). Silicone-induced human adjuvant disease. Ann. Plast. Surg., 10: 270–273. Bohach, G.A., Fast, D.J., Nelson, R.D. and Schlievert, P.M. (1990). Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses. Crit. Rev. Microbiol., 17: 251–272. Bommer, J., Waldher, R. and Ritz, E. (1983). Silicone storage disease in long-term hemodialysis patients. Contr. Nephrol., 36: 115–126. Byrne, J. (1996). Informed Consent. McGraw-Hill, New York. Byrne, K. (1992). Medical records in litigation: The Dalkon shield story, A.M.R.O., 2: 11–14. Carter, J.E., Roll, D.B. and Petersen, R.V. (1974). The in vitro hydrolysis of di(2-ethylhexy)phthalate by rat tissues. Drug Metab. Dispos., 2: 341. Cohen, W.N. and Safaie-Shriazi, S. (1973). Starch granulomatous peritonitis. Am. J. Roentgenol. Radium Ther, Nucl. Med., 117: 334–339. Daniel, J.W. and Bratt, H. (1973). The absorption, metabolism and tissue distribution of di(2-ethyl-hexyl)phthalate in rats. Toxicology, 2: 51. Eckhardt, R.E. and Hindin, R. (1973). The health hazards of plastics. J. Occup. Med., 15: 809. Eismann, B., Seelig, M.G. and Womack, N.A. (1947). Talcum granuloma: Frequent and serious complication. Ann. Surg., 126: 820–832. Estlander, T., Jolanski, R. and Kanerva, L. (1986). Dermatitis and urticaria from rubber and plastic gloves. Contact Dermatitis, 14: 20–25. Fay, M.F. and Sullivan, R.W. (1992). Changing requirements for glove selection and hand protection. Biomed. Instr. Technol., 26: 227–232. Fisher, A.A. (1991). Management of allergic contact dermatitis due to rubber gloves in health and hospital personnel. Cutis, 47: 301–302. Gad, S.C. (1999). Product Safety Assessment. Shared Dehper, New York. Gonzales, E. (1992). Latex hypersensitivity: A new and unexpected problem. Hosp. Pract., 15: 137–151. Guess, W.L. and O’Leary, R.K. (1969). Toxicity of a rubber accelerator. Toxicol. Applied Pharmacol., 14: 221–231. Heggers, J.P., Kassovsky, N., Parsons, R.W., Rosbon, M.C., Pelley, R.P., and Raine, T.J. (1983). Biocompatibility of silicone implants. Ann. Plast. Surg., 11: 38–45. Henderson, W.J., Melville-Jones, C., Barr, W.T., and Griffiths, K. (1975). Identification of talc on surgeons’ gloves and in tissue from starch granulomas. Br. J. Surg., 62: 941–944. Hennepens, C.H., Lee, I., Cook, N.R. et al. (1996). Self-reported breast implants and connective tissue diseases in female health professionals. JAMA, 275: 616–621. Houppert, K. (1995). Embarrassed to death. Village Voice, Feb. 7. Ignatius, J.D. and Hartmann, W.M. (1972). The glove starch peritonitis syndrome. Ann. Surg., 175: 388–397. ISO (2000). Medical Devices – Risk Management–Application of Risk Analysis, ISO 14971-1. ISO (2007). Biological Evaluation of Medical Devices–Development of tolerable intake values for Di-(z-ethylhexyl) phthalate (DEHP) ISO/PDTS 29741. Jaeger, R.J. and Rubin, R.J. (1973). Extraction, localization and metabolism of di-2-ethylhexyl phthalate from PVC plastic medical devices. Environ. Health Perspect., 3: 95. Kelly, K.J., Kurp, V.P., Reijula, K.E., and Fink, J.N. (1994). The diagnosis of natural rubber latex allergy. J. Allergy Clin. Immunol., 93: 813–816. Lake, B.G., Branton, P.J., Gangoli, S.D., Butterworth, K.R. and Grasso, P. (1976). Studies on the effects of orally administered di(2-ethylhexyl) phthalate in the ferret. Toxicology, 6: 341. Lang, L.D. (1996). A review of latex hypersensitivity. Toxic Sub. Mech., 15: 1–11. Laohapand, T., Osman, E.M., Morley, A.R., Ward, M.K., and Kerr, D.N.S. (1982). Accumulation of silicone elastomers in regular dialysis. Proc. EDTA, 19: 143–152.

Chapter 20: Lazarus, J.M. (1980). Complications in hemodialysis: an overview. Kidney Int., 18: 783–796. Lee, C.M., Collins, W.T., and Largen, T.L. (1952). Reappraisal of absorbable glove powder. Surg. Gynecol. Obstet., 95: 725–737. Lee, C.M., and Lehman, E.P. (1947). Experiment with non-irritating glove powder. Surg. Gynec. Obstet., 84: 689. Lichtman, A.L., McDonald, J.R., Dixon, C.F. and Mann, F.C. (1946). Talc granuloma. Surg. Gynecol. Obstet., 83: 531–546. Maguire, D. (1966). Tampon vaginitis. J.S.C. Med. Assoc., 62: 432. Mayer, Jr., F.L., Stalling, D.L. and Johnson, J.L. (1972) Phthalate esters as environmental contaminants. Nature, 238: 411. McAdams, G.B (1956). Granuloma caused by absorbable starch glove powder. Surgery, 39: 329–336. McGregor, R.R. (1953). Silicones in pharmacy. Pharmacy International, Jan. 24–26, 63. Meek, J.H. and Pettit, B.R. (1985). Avoidable accumulation of potentially toxic levels of bevothiazoles in babies receiving intravenous therapy. Lancet, 2: 1090–1092. Miripol, J.E. and Stern, I.J. (1977). Decreased accumulation of phthalate plasticizer during storage of blood as packed cells. Transfusion, 17: 71–72. Mumford. S.D. and Kessel, E. (1992). Was the Dalkon shield a safe and effective intrauterine device? The conflict between casecontrol and clinical trial findings. Fertility and Sterility, 57: 1151–1176. Nash, D.F.E. (1971). Glove powder. Br. Med. J., 3: 329–336. Neely, J and Davies, J.D. (1971) Starch granulomatosis of peritoneum. Br. Med. J., 3: 625–629. Nikonorow, M., Mazur, H. and Piekacy, H (1973). Effect of orally administered plasticizers and polyvinylchloride stabilizers in the rat. Toxic. Appl. Pharmoc., 26: 253–259. Nocera, J. (1995). Fatal litigation. Fortune, Oct 16, pp. 60–82. NRC (1983). Risk Assessment in the Federal Government: Managing the Process. Washington DC, National Research Council. Orem, C. and Beck, J. (1981). The tampon: Investigated and challenged. Women Sc. Health, 6: 105–122. Peck, C.C., Odom, D.G, Friedman, H.I. et al. (1979). Di-2-ethylhexyl phthalate (DEHP), and mono-2-ethylhexyl phthalate (MEHP). Accumulation in whole blood and red cell concentrates. Transfusion, 19: 137–146. Perper, J.A., Pidlaon, A. and Fisher, R.S. (1971). Granulomatous peritonitis induced by rice starch. Am. J. Surg., 122: 812–817 Peterson, M.C., Vine, J., Ashley, J.J., and Nation, R.L. (1981). Leaching of 2-(2-hydroxyethylmercapto-)benzothiazole into contents of disposable syringes. J. Pharm. Sci., 70: 1139–1143. Postlethwaite, R.W., Howard, H.L. and Schanher, W.P. (1949). Comparison of tissue reaction to talc and modified starch glove powder. Surgery, 25: 22. Reingold, A.L., Hargrett, N.T., Shands, K.N. et al. (1982a). Toxic shock syndrome surveillance in the United States, 1980 to 1981. Ann. Intern. Med., 96 (Part 2): 875–880. Reingold, A.L., Hargrett, N.T., Dan, B.B., Shands, K.N., Strickland, B.Y. and Broome, C.V. (1982b). Nonmenstrual toxic shock syndrome. A review of 130 cases. Ann. Intern. Med., 96: 871–874. Rock, G., Secours, V.E., Franklin, C.A., Chu, I. and Villeneuve, D.C. (1978). The accumulation of mono-2-ethylhexylphthalate (MEHP)

Case Histories and Problem Resolution

287

during storage of whole blood and plasma. Transfusion, 18: 553–558. Roster, F. and Gruenwald, H.W. (1986). Contaminants in blood of infants on prolonged intravenous therapy. Lancet, 1: 380–381. Sasakawa, S. and Mitomi, Y. (1978). Di-2-ethylhexylphthalate (DEMP) content of blood or blood components stored in plastic bags. Vox Sansg, 34: 81–86. Scarbeck, K. (1993) Latex: Is it safe? Acad. Gen. Dent., 21: 2–6. Schlievert, P.M. and Blomster, D.A. (1983). Production of staphylococcal pyrogenic exotoxin type C: Influence of physical and chemical factors. J. Infect. Dis., 147: 236–242. Schrock, C.G. (1980). Disease alert. JAMA, 243: 1231. Sheftel, V.O. (1994). Handbook of Toxic Properties of Monomers and Additives. CRC Press, Boca Raton, FL. Shumnes, E. and Darby, T (1984). Contact dermatitis due to endotoxin in irradiated latex gloves. Contact Dermatitis, 10(4): 240–244. Singh, I., Chow, W.L. and Charlani, I.V. (1974) Synovial reaction to glove powder. Clin. Orthopedics, 99: 285. Sivin, I. (1993). Another look at the Dalkon shield: Meta-analysis underscores its problems. Contraception, 48: 1–12. Slater, J.E. (1992). Allergic reactions to natural rubber. Ann. Allergy, 68: 203–209. Snierson, H. and Woo, Z.P. (1955). Starch powder granulomata A report of 2 cases. Ann. Surg., 142: 1045. Sobel, H.J., Schiffman, R.J., Camden, N.J., Schwarz, R. and Albert, W.S. (1971). Granulomas and peritonitis due to starch glove powder. Arch.. Pathol., 91: 559–568. Sobol, R.B. (1991). Bending the Law. The Story of the Dalkon Shield Bankruptcy. University of Chicago Press, Chicago. Stark, N.J. (1996). Managing Positive Biocompatibility Test Results. Medical Device Diagnostic Industry, Oct., pp.148–163. Sussman, G.L. (1995). Allergy to latex rubber. Ann. Intern. Med., 122: 43–44. Sussman, G.L., Taro, S. and Dolovich, J. (1991). The spectrum of IgE-mediated responses to latex. J. Am. Med. Assoc., 265: 2844–2847. Taft, D.A., Lasersohn, J.T., and Hill, L.D. (1970). Glove starch granulomatous peritonitis. Am. J. Surg., 120: 231–236. Tatum, H.J., Schmidt, F.H., Phillips, D., McCarty, M. and O’Leary, W.M. (1975). The Dalkon shield controversy: Structural and bacteriological studies of IUD tails, JAMA, 231: 711–717. Thomas, D. and Withington, P.S. (1985). Toxic shock syndrome: A review of the literature, Ann. R. Coll. Surg., 67: 156–158. Todd, J., Fishaut, M., Kapral, F. and Welch, T. (1978). Toxic shock syndrome associated with phage group 1 staphylococci. Lancet, 1: 16–18. Turjanama, K. (1993). European medical experiences from latex protein allergy: The present position. Proc. Int. Conf. Rubber Consultants and the European Rubber Journal, pp. 17–19. Uber, C.L. and McReynolds, R.A. (1982). Immunotoxicology of silica. CRC Reviews Toxicol., 10: 303–320. Warpinski, J.R., Folgert, J., Cohen, M. and Bush, R.K. (1991). Allergic reaction to latex: A risk factor for unsuspected anaphylaxis. Allergy Proc., 12: 95–102. Wong, O. (1996). A critical assessment of the relationship between silicone breast implants and connective tissue diseases. Reg. Toxicol. Pharmacol., 23: 74–85.

Appendix A Selected Regulatory and Toxicological Acronyms and Abbreviations 510(k) AALAS AAMI ABT ACGIH ACT ADI AIDS AIMD ANSI APHIS ASTM CAS CBER CDER CDRH CFR CFAN CIIT CPMP CPSC CRF CSE CSM CTC CTX CVM DART DHHS DIA DIC DMF DOE DOT DSHEA EEC EM EPA EU FCA FDA FDC FDCA FDLI FHSA FIFRA GCP

Premarket notification for change in a device American Association Laboratory Animal Science Association for the Advancement of Medical Instrumentation American Board of Toxicology American Conference of Governmental Industrial Hygienists American College of Toxicology Allowable Daily Intake Acquired Immune Deficiency Syndrome Active Implantable Medical Device American National Standards Institute Animal and Plant Health Inspection Service American Society for Testing and Materials Chemical Abstract Service Center for Biologic Evaluation and Research (FDA) Center for Drug Evaluation and Research (FDA) Center for Devices and Radiological Health (FDA) Code of Federal Regulations Center for Food and Distribution (FDA) Chemical Industries Institute of Toxicology Committee on Proprietary Medicinal Products (U.K.) Consumer Product Safety Commission Code of Federal Regulations Control Standard Endotoxin Committee on Safety of Medicines (U.K.) Clinical Trial Certificate (U.K.) Clinical Trial Certificate Exemption (U.K.) Center for Veterinary Medicine (FDA) Development and Reproduction Toxicology Department of Health and Human Services Drug Information Associates Disseminated Intramuscular Coagulation Device (or Drug) Master File Department of Energy Department of Transportation Dietrary Supplement Health and Education Act European Economic Community Electron Microscopy Environmental Protection Agency European Union Freund’s Complete Adjuvant Food and Drug Administration Food Drug and Cosmetic Act Food, Drug and Cosmetic Act Food and Drug Law Institute Federal Hazardous Substances Act Federal Insecticides, Fungicides and Rodenticides Act Good Clinical Practices

GLP GMP GPMT HEW HIMA HSDB IARC ICH id IDE IND(A) Ip IRAG IRB IRLG ISO IUD Iv JECFA JMAFF LA LAL LD50 LOEL MAA MD MHW MID MOE MOU MRL MSDS MTD NAS NCTR NDA NIH NIOSH NK NLM NOEL NTP ODE OECD PDI PDN PEL PhRMA

289

Good Laboratory Practices Good Manufacturing Practices Guinea Pig Maximization Test Department of Health, Education and Welfare (no longer existent) Health Industry Manufacturer’s Association Hazardous Substances Data Bank International Agency for Research on Cancer International Conference on Harmonization Intradermal Investigational Device Exemption Investigational New Drug Application Intraperitoneal Interagency Regulatory Alternatives Group Institutional Review Board Interagency Regulatory Liaison Group International Standards Organization Intrauterine Device Intravenous Joint Expert Committee for Food Additives Japanese Ministry of Agriculture, Forestry, and Fishery Licensing Authority (U.K.) Limulus amebocyte lysate Lethal dose 50: The dose calculated to kill 50% of a subject population, median lethal dose Lowest observed effect level Marketing Authorization Application (EEC) Medical device Ministry of Health and Welfare (Japan) Maximum implantable dose Margin of Exposure Memorandum of Understanding Maximum Residue Limits Material Safety Data Sheet Maximum tolerated dose National Academy of Science National Center for Toxicological Research New Drug Application National Institutes of Health National Institute Occupational Safety and Health Natural Killer National Library of Medicine No-Observable-Effect Level National Toxicology Program Office of Device Evaluation Organization for Economic Cooperation and Development Primary Dermal Irritancy Product Development Notification Permissible Exposure Limit Pharmaceutical Research and Manufacturers Association

290 PL PLA PMA PMOA PMN po PTC QAU RAC RCRA RTECS SARA

Appendix A

Produce License (U.K.) Produce License Application Premarket Approval Applications Principal Mode of Action Premanufacturing Notice Per os (orally) Points to Consider Quality Assurance Unit Recombinant DNA Advisory Committee Resources Conservation and Recovery Act Registry of Toxic Effects of Chemical Substances Superfund/Amendments and Reauthorization Act

sc SCE SNUR SOP SOT SRM STEL TLV TSCA USAN USDA USEPA USP WHO

Subcutaneous Sister Chromatic Exchange Significant New Use Regulations Standard Operating Procedure Society of Toxicology Standard Reference Materials (Japan) Short Term Exposure Limit Threshold Limit Value Toxic Substances Control Act United States Adopted Name Council United States Department of Agriculture United States Environmental Protection Agency United States Pharmacopoeia World Health Organization

Appendix B Contract Testing Laboratories

Laboratory

Address

Phone #

Accel Lab

1635 Lionel-Bertrand Boisbriand (Quebec) J7H 1N8

(450) 435-9482

American Pre-Clinical Services, LLC

Minneapolis, MN 55433

(877) 717-7997

AppTec (Wuxi) (3 sites)

2540 Executive Drive St. Paul, MN 55120

651-675-2000

Consumer Product Testing Company

70 New Dutch Lane Fairfield, NJ 07004

(973) 808-7111

CV Path Insitute, Inc.

19 Firstfield Road Gaithersburg, MD 20878

(301) 208-3570

Lychron, LLC

2569 Wyandotte Street Mountain View, CA 94043-2314

(650) 938-3675

MB Research Laboratories, Inc.

P.O. Box 178 Spinnertown, PA 18968

(215) 536-4110

MicroTest Laboratories, Inc.

P.O. Box 848 Agawam, MA 01001

(413) 789-4334

Nelson Laboratories

6280 South Redwood Road Salt Lake City, UT 841123-6600

800-826-2088

North American Science Associates, Inc. (4 sites)

2261 Tracy Road Toledo, OH 43619

(419) 227-6882

Northview Pacific Laboratories, Inc.

551 Linus Pauling Dr. Hercules, CA 94547

(510) 964-9000

SGS U.S. Testing Company, Inc.

291 Fairfield Avenue Fairfield, NJ 07004

(201) 575-5252

STS duoTEK, Inc.

7500 W. Henrietta Road Rush, NY 14543

(716) 533-1672

SURPASS

803 Prospect Avenue Osceola, WI 54020

(715) 294-4371

TOXIKON

15 Wiggins Avenue Bedford, MA 01730

(617) 275-3330

Viromed Biosafety Laboratories

6101 Blue Circle Drive Minneapolis, MN 55343

(800) 582-0077 (952) 563-3926

291

Appendix C Notable International Regulatory Internet Addresses

Organization or Publication

Web Address (URL)

Sample Main Topics

Advanced Medical Technology Association (AdvaMed), formerly known as HIMA Agency for Toxic Substances and Disease Registry Australian Therapeutic Goods Administration

http://www.advamed.org/memberportal/

About AdvaMed, Issues and Advocacy, Meetings and Events, Publications, Business Solutions

ChemInfo Code of Federal Regulations

www.atsdr.cdc.gov http://www.tga.gov.au/

Cornell Legal Library

www.indiana.edu/~cheminfo/ca_csti.html http://www.access.gpo.gov/nara/cfr/ cfr-table-search.html http://www.law.cornell.edu

EPA European Medical News European Medicines Agency Food and Drug Administration (FDA)

www.epa.gov http://www.eucomed.com/ http://www.emea.europa.eu/ www.fda.gov

FDA—CDRH Comment Device Advice PDF Reader General Principles of Software Validation (draft) Guidance for Industry Division of Small Manufacturers Assistance

http://www.fda.gov/cdrh/ www.fda.gov/cdrh/comment4.html www.fda.gov/cdrh/devadvice/32.html www.fda.gov/cdrh/acrobat.html http://www.fda.gov/cdrh/comp/guidance/ 938.html http://www.fda.gov/cdrh/industry/support/ www.fda.gov/cdrh/fr1007ap.pdf www.fda.gov/cdrh/gmpasci.zip www.fda.gov/cdrh/comp/designgd.html

Design Control Guidance for Medical Device Manufacturers Design Control Inspection Results Do It By Design: An Introduction to Human Factors in Medical Devices Human Factors Implications of the New GMP Rule Medical Device Quality Systems Manual: A Small Entity Compliance Guide Guidance on Information Disclosure by Manufacturers to Assemblers for Diagnostic X-ray Systems

Guidance on Quality System Regulation Information for Various PreMarket Submissions text

CDRH Letter to Manufacturers

GMP/QS Workshops with CDRH Participation Quality System Inspections Reengineering FDA—Field Operations

www.fda.gov/cdrh/comp/designgd.pdf www.fda.gov/cdrh/dsma/dcisresults.html www.fda.gov/cdrh/humfac/doit.html www.fda.gov/cdrh/humfac/doitpdf.pdf www.fda.gov/cdrh/humfac/hufacimp.html http://www.fda.gov/cdrh/dsma/ gmpman.html http://www.fda.gov/cdrh/comp/guidance/ 2619.html http://www.fda.gov/cdrh/comp/guidance/ 2619.pdf http://www.fda.gov/cdrh/comp/guidance/ 1140.html http://www.fda.gov/cdrh/comp/guidance/ 1140.pdf www.fda.gov/cdrh/yr2000/cdrh/letters/ 980921/y2kcompltr.html www.fda.gov/cdrh/yr2000/cdrh/letters/ 980921/y2kcompltr.pdf www.fda.gov/cdrh/dsma/workshop.html www.fda.gov/cdrh//gmp/gmp.html www.fda.gov/ora/

293

Medical Devices; GMP Codes; Parliamentary Secretary’s Working; Status Document; Party on Complementary Medicines; Medical Releases; Publications; Site Map; Related Sites SirCH: Chemical Safety Or Toxicology Information NARA Code Sections Code of Federal Regulations; Supreme Court Decisions; U.S. Code; Circuit Courts of Appeal News; Related Sites What’s New; Documents Forum; Other Sites Foods; Human Drugs; Biologics; Animal Drugs; Cosmetics; Medical Devices/Radiological Health Home Page Questions and Feedback on CDRH Site

Text Good Manufacturing Practice (GMP) (also known as the Quality System Regulation) Final Rule Text as Published in the Federal Register Text PDF version First Year Roll Out Final Design Control Inspection Results and Presentation Text By Dick Sawyer Office of Health and Industry Programs PDF version Text Text

PDF version

PDF version Letter to Manufacturers—Y2K Issue for Production Processes and Quality System Software Text PDF version

What’s New; Import Program; Inspectional, Science and Compliance References; Federal/State Relations

294

Appendix C

Organization or Publication

Web Address (URL)

Sample Main Topics

Design Controls

http://www.fda.gov/cdrh/comp/ designgd.html http://www.fda.gov/cdrh/comp/ designgd.pdf http://www.fda.gov/ora/inspect_ref/igs/ elec_med_dev/emc1.html

Design Control Guidance For Medical Device Manufacturers Text PDF version

Guide to Inspections of Quality Systems

Food and Drug Law Institute

www.fda.gov/ora/inspect_ref/igs/qsit/ qsitguide.htm www.fda.gov/ora/inspect_ref/igs/qsit/ QSITGUIDE.PDF http://www.fdli.org

Health Canada

http://www.hc-sc.gc.ca/index_e.html

Japanese Ministry of Health and Welfare

http://www.mhlw.go.jp/english/index.html

Library of Congress

http://thomas.loc.gov

Medical Device Link

http://www.devicelink.com

Medical Device Manufacturers Association (MDMA)

http://www.medicaldevices.org/public/

National Archives and Public Records Administration

http://www.archives.gov/

National Library Network

www.toxnet.nlm.nih.gov

National Toxicology Program New Quality System (QS) Regulation Regulatory Affairs Professionals Society (RAPS)

http://ntp-server.niehs.nih.gov/ www.fda.gov/bbs/topics/ANSWERS/ ANS00763.html http://www.raps.org

U.S. Department of Commerce

http://www.commerce.gov/

University of Pittsburgh World Health Organization

www.pitt.edu http://www.who.int/en/

Guide to Inspections of Electromagnetic Compatibility Aspects of Medical Device Quality Systems Text QSIT Inspection Handbook Text PDF version Special Interest; Publications; Multimedia; Order Products; Academic Programs; Directory of Lawyers and Consultants; Contact Us Latest Headlines, Advisories/Warnings/Recalls, Consumer Product Safety, Diseases and Conditions, Drugs and Health Products Organization; Y2K Problem; Statistics; White Paper; Related Sites Searchable Database of Federal Legislation, Congressional Record and Committee Information News; Consultants; Bookstore; Links; Discussion; Magazines (MDDI; MPMN; IVD Technology) About MDMA; Issues; Programs; Advocacy; Resources Code of Federal Regulations; Federal Register; Laws; U.S. Congress Information TOXNET: Toxicology Data Network, a Cluster of Databases on Toxicology, Hazardous Chemicals, and Related Areas FDA Talk Paper Announcing the GMP Final Rule Text Certificates; Resource Center; Publications; Chapters; Related Links; Contacting RAPS Bureau of Industry and Security; International Trade Association; Patent and Trademark; National Institute of Standards and Technology Governance; Health Topics; Information Sources; Reports; Director-General; About WHO; International Digest of Health; Legislation

Appendix D Non-U.S. Medical Device Regulators

Organization or Publication

Web Address (URL)

Sample Main Topics

European Union (EU)

http://europa.eu.int/

Health Canada

www.hc-sc.gc.ca

Medical Device Agency

http://www.mhra.gov.uk/home/ idcplg?IdcService=SS_GET_PAGE&nodeId=5 www.nihs.go.jp

This is a multi-lingual gateway to information about all activities of the European Union Provides useful information about the regulation of all medical products in Canada, and other Canadian government health programs Provides useful information about the regulation of medical devices in the United Kingdom Provides useful information about the regulation of medical devices and pharmaceutical products in Japan Provides useful information about regulation and safety of medical devices and pharmaceutical production in Switzerland Chinese FDA Provides useful information about the regulation of medical devices and pharmaceutical products in Australia

National Institute of Health Sciences—Japan SWISSMFDTC SFDA Therapeutic Goods Administration

www.swissmedia.ch www.sfda.gov.cn http://www.tga.gov.au/

295

Index

Absorption, 259–60 Active transport, 258 Active/nonactive devices, combined characteristics, 25 Acute irritation, see primary irritation Acute toxicity testing, see toxicity Ad libritum feeding, 198 Adduct formation, 156–7 ADME process, 255–6, 273 Agar diffusion test, 72–3 Agar overlay tissue culture, 72 Agarose, 71 AIDS, 128, 279 Air, 264, 267 Ames Salmonella/Microsome Test, 160 Amino acids, 151, 162 Anaphylaxis test, 113 Aneuploidy, 159 Angiography, 80 Animal Care and Use Committees, 181 Animal Welfare Act 1989 (AWA), 4, 181 Annual Plan for Fiscal Year 1996 (NTP), 42 ANSI/AAMI ST60 (1996), 211 ANSI/AAMI/ISO standards, 19 Anti-infective materials, 244 Antibodies, 101 Antigen-presenting cells (APCs), 100–2 ASTM Standard F 746, 205 ASTM Standard F 2129, 205 ASTM standard practice, 139, 141 Auto-immunity, 105–6, 128 Autoantibodies, 106 Autoradiography, 271–2 B-cells, 106, 110 Bacteria endotoxins 30 mutation tests 162 Bile, 267 Bioactive metals, 50–1 Biocompatibility, 1–2, 11, 28–9, 32, 69, 280–2 Biologic Evaluation and Research Center (CBER), 8, 224–5 Biological tests, 13 Biologically derived materials, 28, 38, 49 Bioresearch monitoring program (BIMO), 231 Blinding see masking Blood, 266–71, 275 Body weight, 147, 183 Bone, 51, 137–8 Breast implants, 280–1 British pharmacopoeia, 133–6 British Standard 5736/2 (1981), 133 Brown-Norway strains, 113 Buehler tests, 115–18 Bureau of Medical Devices and Diagnostic Products, 3 Calcification, 209 Canadian Centre for Occupational Health and Safety (CCINFO), 44 Cancer, 157–62, 166, 168, 197–201 Cannulae, 244 Carbon-based nanomaterials, 59 Card assignment, 64 Cardiovascular devices, 243–5

Catheters/guidewires, 244 CDRH Contact Lens Branch 245 immunotoxicity testing matrix 110 Obstetrics-Gynecology Devices Branch 245 testing framework 108, 235 Cell-free/cell-based systems, 161 Cell-mediated cytotoxicity, 101 Cell-mediated immunity, 11 Cells, 172, 174 antigen-presenting (APCs) 100–2 death 69 function tests 71 growth assay 71 immunity 103 lines 69–70 Center for Biologic Evaluation and Research (CBER), 8, 224–5 Center for Devices and Radiological Health (CDRH), 5, 8–9, 224–8 Center for Drug Evaluation and Research (CDER), 8, 224–9 Center for Food Safety and Applied Nutrition, 99 Centric fusions, 159 Ceramics/glass, 35–9, 49, 51–2, 206 Chain scission processes, 213–14 CHEM, 59 Chemical hazards classification, 40 Chemical sterilization, 29, 212 Chemotherapy, 103, see also cancer Chi-square test, 64, 200 Chromosomal aberrations, see elastogenicity Chromosomes, 151, 159–60, 174–5 Circumcision kits, 285 Claims, 39 Class I products, 7, 16, 21 Class II products, 7 Class III products, 7 Clastogens, 172–4 Clinical chemistry, 108–9 Clinical pathology, 183–4 Cloning efficiency assay, 71 “Closures” (plastics), 69 Clotting time, 81 Cluster sampling, 63 Coagulation, 81 Cobalt-based alloys, 49 Code of Federal Regulations (21 CFR), 7, 236 Cohort studies, 238–40 College of Obstetrics and Gynecology (ACOG), 246 Colorimetric readings, 92 Combination products, 223–9, 232 Committee for Clinical Review, Final Report, see Temple Report Common biocompatibility assays, 45 Complement activation, 80–1 Complement system, 110 Computerized algorithm, 64 Concanavalin A (Con A), 111 “Constitution”, 52 Consumer Product Safety Commission, 83, 126 Contact lenses, 245

297

Control standard endotoxin (CSE), 92–3 Cooper Committee see Medical Devices Study Group, 1969 Copolymers, 33, 52, 55–6, 208–10 Cornstarch, 277 Corrosion crevice 204 fretting 204 galvanic 205 stress 204–5 testing 205 Covariate adjustment, 236 CRC Critical Reviews in Toxicology, 41 Crystal violet staining, 70 Cumulative irritation, 83 Current contents, 42 Cyclophosphamide, 102 Cytogenetics, 158–9, 171–2 Cytotoxic T lymphocyte (CTL) mediated assay, 111 Cytotoxicity testing, 14, 69–73, 77, 102, 107, 167–8, 173, 175, 205, 285 Dacron, 212–13 Dalkon Shield, 4, 276–7 Deaths, 147, 238–9 Degradation, 29, 203–20 DEHP, 57, 210–11, 275–6 Delayed-type hypersensitivity (DTH), 111, 121, 206 Dendrimers, 60 Department of Health and Human Services (DHHS), 8 Dermal irritation, 83–5 Descriptive studies, 240 Device categories, 13 Device Experience Network (DEN), 278 Device master file (DMF), 21, 45 Device/material implantation effects, 16 Devices and Radiological Health Center (CDRH), 5, 8–9, 224–8 Diet, 188 Diethylhexylphthalate (DEHP), 32, 54 Dimethylnitrosamine, 102 Diphtheria, 105 Direct contact, 71–3 DNA, 151, 153–62, 166, 171, 175 Dogs, 137–40, 182, 244 Dose-response assessment, 22, 284–5 Double-blind trials, 233–5 Draize rabbit skin test, 83, 86, 115 Drug Evaluation and Research Center (CDER), 8, 224–9 Dry heat 14 E. coli, 154, 162–3 Echocardiography, 245 ECVAM (Europe), 86 Edema, 84–5 Elastogenicity, 151 Elastomers, 32–3, 37, 53–7 Elution test, 73 Embryo-fetal development, 187 Embryotoxicity, 245 End points, 234 Endotoxin, 92–4

298

Index

Environmental Protection Agency (EPA-USA), 30, 191, 237 Environmental stress cracking (ESC), 209 Enzyme-linked immunosorbent assays (ELISAs), 109, 206 EPA Gene-Tox Program (USA), 168 Epidemiology, 37, 238 Epidermis, 84 Epigenetic mechanism, 197 Erythremia, 84–5 Erythrocyte stability, 80 Ethylene oxide, 212–13 Ethyleneterephthalate (PET), 212 Eukaryotic genes, 151, 153, 159, 166 Eukaryotic mutation tests, 166 European Economic Community, 14–18 European standard EN 980, 21 EVCAM/IACVAM, 88 Excision repair pathway, 153 Excretion, 264, 267, 269 Exposure assessment, 22, 284–5 External communicating devices, 75–6 Extra vascular hemolysis, 78 Extract dilution, 71 Extractants and vehicles, 65–7 Eye irritation, 14, 86, 250–1 Facilitative transport, 258 “Factors of influence”, 38–9 FDA device categories/ suggested biological testing, 12 Feces see excretion Federal Food, Drug and Cosmetic Act, 1938 7 Fertility, 186–7 Fetal examinations, 189 Fibers, 57–8 Fibrin/fibrinogen degradation products, 81 Fibroblasts, 70, 72 Fibrosarcomas, 197–8 Fisher’s exact tests, 200 Fluid medium tissue culture, 71–2 Food consumption, 183 Food and Drug Act 1906, 3 Food and Drug Administration (FDA), 3, 5, 7–8, 12, 15, 21, 181, 245, 251, 279 Food and Drug Law Institute (FDLI), 5, 8 Food Safety and Applied Nutrition Center, 99 Foreign antigens, 99 Formulation additives, 29 Forward mutations tests, 162, 166 Fruit flies, 166 FTIR analysis, 218 Functional reserve capacity, 128 Gamma irradiation, 29 GC/MS semi-volatile organic compounds, 218 GC/MS-volatile organic compounds, 217–18 Gel diffusion, 71 Genetic damage, 157–8 Genitalia, 189–90 Genotoxicity, 16, 151–74 Germ cell assays, 174 Goats, 137–8 Good clinical practice regulations (GCP), 235 Good laboratory practices (GLPs), 181, 237 Gower Publishing (Brookfield, VT), 44 Gravimetric analysis, 80 Guide for the Care and Use of Laboratory Animals, 140 Guinea Pig Maximization Test (GPMT), 118–21 Guinea pigs, 115–20, 124–5, 136

Habermann, V., 212 Hamsters, 89, 161, 167–8, 172, 182 Hank’s solution, 204 Haptens, 103 Hazard identification, 21–2, 39, 40, 284–5 Health care providers, 45 Heart valve prostheses, 245 Hematology, 108 Hemocompatibility, 75–82 Hemolysis tests, 15, 77 Hepatocytes, 161 Heritable chromosome assays, 174 Histopathology, 109, 184–6 Homopolymers, 31–2, 52–4 Horny layer, 84 Host-resistance assays, 113 Human peripheral blood lymphocytes, 172 Humoral immunity, 101, 103, 109 Hydrogels, 58 Hydrolysis, 208 Hydrophilicity, 208 Hypersensitivity, 99–100, 103–5, 113, 127–8, 279–80 Hypospadias, 190 ICH test profile, 160 ICP-metal content, 218 Immune system, 99–103, 107 Immunopathology, 100 Immunoregulatory circuit, 101 Immunosorbent assay methods (ELISA), 206 Immunostimulation, 102–3, 107 Immunosuppression, 102–3, 107, 128 Immunotoxicology, 99–128 Implant devices, 51, 76, 80 Implantation, 133–42, 205, 216–17, 280–1 In vitro fertilization (IVF), 206, 209, 245, 270 Inadvertent contaminants, 29 Indirect immunotoxic effects, 127 Inflammatory responses, 252–3 Informed consent (Byrne), 280 Inorganic nanomaterials, 59–60 Institutional Animal Care and Use Committee (IACU), 4 Institutional review boards (IRB), 236 Interagency regulatory liaison group (IRLG), 147 Intercenter agreements, 224–5 Interim analyses, 236 International Agency for Research on Cancer, 197 International Conference on Harmonization (ICH), 186–7 International Organization for Standardization (ISO), 282 International System Human Cytogenetic Nomenclature, 173 International Veterinary pathology slide bank (IVPSB), 44 Intracutaneous irritation, 91 Intracutaneous test, 14, 66 Intramuscular injection (IM), 91 Intramuscular irritation, 90 Intrauterine devices (IUD), 276–7 Intravenous irritation, 90 Intravenous injection (IV), 91 Investigational Device Exemption (IDE), 235–6 Irritation studies/tests, see local tissue tolerance Irritation/toxicity screen, 114–15 ISO 1137, 213 ISO 10993, 134–5, 160, 211, 216 ISO 10993-1, 282 ISO 10993-3, 151

ISO 10993-4, 75, 82 ISO 10993-6, 133 ISO 10993-10, 2, 13–14 ISO 10993-11, 197 ISO 10993-12, 63, 67 ISO 10993-14, 30 ISO 10993-20, 99 ISO 11134, 212 ISO 11135, 212 ISO 14538, 284 ISO 14971-1, 283 ISO evaluation tests, 14–15, 18–19 Japanese ISO test selection guidelines, 15 Japanese MHW test selection guidelines, 20 Karber’s formula, 148 Kawasaki’s disease, 278 LAL reagent sensitivity, 93–4 Latex allergy, 279–80 Leo White test, 80 Leukemia, 198 Ligature material, 23, 285 Light microscopy method 80 Limulus amebocyte lysate test, 92 Litchfield-Wilcoxon method, 148 Local lymph node assay (LLNA), 124–5 Local tissue tolerance, 83–96 Lymphocyte transformation testing (LTT), 206 Lymphocytes, 100–1 Lymphoid tissues, 109 Macromolecules, see polymers Macrophage function, 112 Magnesium, 51 Mammalian Mutation Tests, 166–7, 170–1 Masking, 234–5 Mast cells, 104, 113 Material safety data sheets (MSDS), 43 Materials, 49–60 Maximum implantable dose (MID), 182, 198 Maximum valid dilution (MVD), 94 Mechanical failure, 28–9 Mechanical insult, 28–9 MEDDEV 10/93 (paper), 16 Medical, Biologic Effects of Environment Pollutants (NAS), 41 Medical Device Amendment Act 1976, 8 Medical Device Classification Flowchart, 9–11 Medical Device Paradigm, 284 Medical Device Safety Act 1967, 3 Medical Devices Classification Overview, 24 Medical Devices Study Group 1969, 3 Medline data base, 44 Melanocyte system, 84 Merck Index, 42 “Metabolic co-operation”, 167 Metabolism/biotransformation, 261–3, 269–71 Metal hypersensitivity, 206 Metallic corrosion, 203–6 Metals, 30–1, 49, 215–16 MHW Guidelines ,1995 13–14 Mice, 121–4, 127–8, 137, 168–9, 171, 174, 187, 245–51, 271 Micronuclei, 174 Microtox test, 70 Milk, 267–8 Mixed lymphocyte response assay (MLR), 111 Monomers, 53, 55–6, 211 Mortality adjusted methods, 200

Index Mouse Ear Swelling Test (MEST), 121–4 Mouse embryo assays, 246–50, 247 Mouse lymphoma l5178y tk test, 168–9 Mouse safety test, 66, 243 Mouse specific locus test, 171 mRNA molecules, 153, 167 Mucosal irritation, 89 Muscles implantation 137–40 irritation 90, 90 Mutagenicity 151, 158–61 Mycobacterium tuberculosis, 112 Nanomaterials, 59 National Academy of Sciences (USA), 284 National Library of Medicine (NLM), 41–2 National Toxicology Program (NTP-USA), 107, 127, 157 Natural killer cell assays, 112 NCTR ED01 study, 199 Necrosis, 84 Neoplasia, 158–9 Nernst equation, 203 Neuromuscular screen observations, 147 Neutral red uptake assay, 70–2 Nickel, 283 Nitinol, 22, 50, 284 No adverse effect level (doses), 182 No-observable-adverse-effect level (Noael), 283 Non-contact devices, 75 Nonabsorbable suture materials, 58 Nonspecific immunity, 112 Normit-chi square (Berkson), 148 Norplant contraceptive, 281–2 NOSTASOT method, 192 Obstetrics and Gynecology College (ACOG), 246 Ocular irritation, see eye irritation Odds ratio (OR), 240 “Oppenheimer Effect”, 197 Organ/body weights, 108 Osmotic fragility test, 78–9 Oxidation, 209 Parenchymal tumors, 197 Parenteral irritation/tolerance, 89 Partial thromboplastin time (PTT), 81 PASS validation, 149 Passive transport, 258 Passive-pitting-repassivation (PPR), 205 PC-based information products (laser disk), 44 PC-based information products (toxicology), 43 Perchloroethylene solvent, 23, 285 Percutaneous transluminal coronary angioplasty device, 231–2 PET fabrics, 214 Peto/International Agency, 200 Pharmacokinetics, 264–5, 272–3 Pharmacopoeial testing, 14 “Phase 0” clinical trials, 143 Phthalate plasticizers, 275–6 Pigs, 137–8, 244 Pitting, 204 Plasma fibrinogen, 81 Plasmid pKM101, 163 Plasmids, 154 Plasticizers, 210 Plastics, 57 Platelets, 81 PMOA (principal mode of action), 223 Point-mutations, 154–5

Polar/nonpolar solvents, 64 Polarization, 203–4 Pollution and Toxicology (POLTOX), 44 Polyamides, 56 Polyester, 55 Polyether urethane (PEU), 208–9 Polyglycolic acid, 29 Polymeric/ceramic control materials, 138 Polymers, 29, 30–2, 37–8, 49, 52–4, 57–9, 208–13, 218 Polystyrene, 33, 53 Polyvinyl chloride (PVC), 212 Practice F361 (1972), 141 Practice F469 (1978), 141 Pre-/postnatal development, 187 Pre-IDE biocompatibility tests, 8 Prefilled syringes, 215 Premarket Approval (PMA), 7, 231 Primary dermal irritation, 83–5 Primary eye irritation test, 86–8 Proctor and Gamble, 278 Protein adsorption, 81 Protein binding, 261 Prothrombin, 81 Pyrogenicity testing, 15, 92 Quality Assurance Unit (QAU), 4 Quantum dots, 60 Rabbit eye irritation test, 70 Rabbit pyrogen test, 94 Rabbits, 83–95, 133, 136–41, 182, 187, 209, 245, 250–1 Radiation, 213–14, 265–6 Radioimmunoassay (RIA), 265 Randomization, 63–4, 233 Rats, 91, 125, 127, 136–40, 161, 174, 182, 186–8, 190–2, 197–8, 267–8 Raw material characterization, 28 Reasonable man concept, 45 Red blood cell hemolysis assay, 91 Redbook (FDA 1993), 99, 110–13, 199 Reference standard endotoxin (RSE), 92 Relative risk (RR) 240 Repeat insult patch test (RIPT), 237 Residual monomers, 29, 32 Residual solvents, 29 Riegel -v- Medtronics, 282 Ringer’s solution 204 Risk Analysis 1993 (CEN standard), 21, 284 Risk assessment 21, 22, 39, 283–5 Risk assessment immunotoxicity flow chart, 108 RNA, 153–4, 167 Robertsonian translocations, see centric fusions Rubber, 33, 55 S9 mix, 162 Safe Medical Devices Act 1990, 4, 223–4 Safety margins, 286 Sampling, 63–8 Scanning electron microscopy, 80 Scleroderma, 281 Screens, 39 SCT/AET principles, 219 Sensitization test procedures, 13, 114 Sex, 148, 187–8, 191 Shape memory polymers (SMPS), 58 Sheep, 137–8 Shintani, S., 89 Short-term exposure limits (STELS), 23 Silicone microphysiometer, 70 Silicone-filled breast implants, 4 Silicones, 33–7, 56–8, 280–2 Silver, 244

299

Simulated body fluid (SBF), 210 Sister chromatid exchange assays (SCES), 174–5 Skeletal fetal examinations, 190 Skeletal muscle cell line, 91 Skin , 86, 116, 126, 206, 259 Skin irritation, see dermal irritation SMART assay test, 166 “Solid-state carcinogenesis”, 198 Sprague-Dawley rats, 197 Stainless steel, 49, 50 Standard tests, 77 Standardized mortality ratio (SMR), 239 Stanford Research Institute, 205 Steam 212 Sterility assurance level (SAL), 211 Sterilization, 45–6, 65, 211–15 Stratified sampling, 63 Stratum corneum, see horny layer Studies animal 240–1 case-control 239–40 clinical 231–41 clinical safety 237 cross-sectional 240 intervention 240 nested case-control 240 proportional mortality 239 prospective cohort 239 special 243–53 Study contact, 235 Study data analysis, 235–6 Subcutaneous tissue implants, 136–7 Suppressor mutations,155–6 Supramid, 276 Suture materials, 58 Synthetic fibers, 35 Synthetic rubbers, 56 Syringes, 27 Systematic sampling, 63 Systemic Injection Test, 66, 251 Systemic injection test, see mouse safety test T-cell lymphoproliferation response, 111 Talcum powder particles, 277 Tampons, 245, 251–2, 278–9 Tantalum, 51 Target of Organ Toxicity (Taylor & Francis), 41 Temple Report, 231–2, 235 Teratogens, 158 Terpolymers, 56 Test animals grading scales, 15 Test procedures, 125, 282–3 Thermoplastics, 53 Threshold limit values (TLVS), 23 Thrombin time (TT), 81 Thrombogenicity, 80 Thromboplastin, 81 Tissue engineering scaffolds, 59 Tissue irritation studies, 84 Titanium, 50 Tolerance, 89 Toxic release inventory (TRI), 43 Toxic shock syndrome (TSS), 251–2, 277–9 Toxic Substance Control Act 1976, 149 Toxic Substance Control Act 1986, 147 Toxic Substances Control Act 1979, 116 Toxicity case studies 283 maternal 193–4 reproductive/developmental 186–94 risk assessment 39–43, 40 subchronic/Chronic Toxicity 181–6 testing 2, 11–13, 143–9, 271

300

Index

Toxicokinetics, 255–73, 256–8, 260–6, 268 Toxicologist, 42 Toxicology, 63, 237, 255 Toxicology societies, 42 Transfer-popliteal lymph node assay, 128 Tribocorrosion, 205 Triple-blind studies, 235 tRNA molecules, 153–4, 162 Tumorigenicity, see cancer Turbidimetric assays, 92 Type I hypersensitivity, 104, 113 Type I and II errors, 200, 234 Type II hypersensitivity, 104 Type III hypersensitivity, 105 Type IV delayed-type hypersensitivity, 105 Type IV hypersensitivity, 113–14 Types II and III hypersensitivity, 113

UHMWPE, 210 UKEMS, 171–3 Uncertainty factors, 26, 285–6 Urinalysis parameters, 200 Urine, 264 USP (1996), 66 USP implantation test (2007), 133–4 USP pyrogen test, 92 USP Sterilization and Sterility Assurance, 215 USP VI (1893), 216 USP VIII (1905), 216 USP XII (1942), 216

Vascular irritation, 89 Ventricular-assist devices, 244–5 Vinyl acetate, 55 Visceral fetal examinations, 190 Viscosity, 38 Vivo standard, 69 Water, 30, 261 whole blood clotting time, 80 Wound-dressing formulation, 22, 284–5 Xenobiotics, 100, 103 Zirconia polycrystal, 206–7

V79 system, 167 Vaginas, 252, 276–8 Vascular grafts, 245

about the book… Capturing the growth of the global medical device market in recent years, this practical new guide is essential for all who are responsible for ensuring safety in the use and manufacture of medical devices. It has been extensively updated to reflect significant advances, incorporating combination products and helpful case examples of current real-life problems in the field. The Third Edition explores these key current trends: • global device markets • continually advancing technology • the increasing harmonization of device safety regulation worldwide Each aspect of safety evaluation is considered in terms of International Standards Organization (ISO), US Food and Drug Administration (FDA), European Union (EU), and Japanese Ministry of Health and Welfare (MHW) perspectives. In addition, the book reflects the role of the continuing growth of technology in the incorporation of science, particularly in the areas of immunotoxicology and toxicokinetics. about the authors... SHAYNE C. GAD is the principal of Gad Consulting Services, Cary, North Carolina, USA. Gad Consulting Services is a 15-year-old consulting firm serving more than 350 clients (including 120 pharmaceutical companies in the US and 35 overseas). Dr. Gad received his B.S. in chemistry and biology from Whittier College, Whittier, California, and his Ph.D. in pharmacology and toxicology from the University of Texas at Austin, Austin, Texas, USA. He has published 39 books and more than 350 chapters, articles, and abstracts in the fields of toxicology, statistics, pharmacology, drug development, and safety assessment. Dr. Gad has more than 30 years of broad-based experience in toxicology, drug and device development, statistics, and risk assessment with specific expertise in neurotoxicology, in vitro methods, cardiovascular toxicology, inhalation toxicology, immunotoxicology, and genotoxicology. He was the past president of the American College of Toxicology and the Roundtable of Toxicology Consultants. Dr. Gad has consulted for the US Food and Drug Administration, Environmental Protection Agency, and the National Institutes of Health. MARIAN G. MCCORD is Associate Professor with appointments in the Department of Textile Engineering, Chemistry, and Science at North Carolina State University, Raleigh, NC, USA, the Joint UNC/NCSU Department of Biomedical Engineering in Raleigh and Chapel Hill, NC, USA, and an Adjunct Associate Professor of Medicine at the University of North Carolina at Chapel Hill. She serves as co-Director of the Atmospheric Plasma Laboratory at the College of Textiles at NCSU, and is a leader in development of Advanced Medical Technology at NCSU (www.ncsu.edu/amt). Dr. McCord received her M.S. in Bioengineering and Ph.D. in Textiles and Polymer Science from Clemson University, Clemson, South Carolina, USA. She has been active in several professional organizations, including the AAMI Protective Barriers Committee, ASTM F23, the Biomedical Engineering Society, the American Society for Engineering Education, and the Society for Biomaterials. Dr. McCord is the recipient of the 2007 Richard Felder Award for Outstanding Service in Teaching and Learning. She has been an invited lecturer and presenter at numerous consortia, meetings, and conferences, and has been a valued mentor and advisor for her students. Printed in the United States of America

•

Gad

McCord

71645

Safety Evaluation in the Development of Medical Devices and Combination Products, Third Edition

Toxicology

Safety Evaluation in the Development of Medical Devices and Combination Products Third Edition

Shayne C. Gad Marian G. McCord